Datasets:

---

chungimungi

/

arxiv-hard-negatives-bm25

like 0

[ " We consider integrals of spherical harmonics with Fourier exponents on the\nsphere $S^n ,\\, n \\geq 1$. Such transforms arise in the framework of the theory\nof weighted Radon transforms and vector diffraction in electromagnetic fields\ntheory. We give analytic formulas for these integrals, which are exact up to\nmultiplicative constants. These constants depend on choice of basis on the\nsphere. In addition, we find these constants explicitly for the class of\nharmonics arising in the framework of the theory of weighted Radon transforms.\nWe also suggest formulas for finding these constants for the general case.\n", " We study the asymptotic clustering (phase-locking) dynamics for the Kuramoto\nmodel. For the analysis of emergent asymptotic patterns in the Kuramoto flow,\nwe introduce the pathwise critical coupling strength which yields a sharp\ntransition from partial phase-locking to complete phase-locking, and provide\nnontrivial upper bounds for the pathwise critical coupling strength. Numerical\nsimulations suggest that multi- and mono-clusters can emerge asymptotically in\nthe Kuramoto flow depending on the relative magnitude of the coupling strength\ncompared to the sizes of natural frequencies. However, theoretical and rigorous\nanalysis for such phase-locking dynamics of the Kuramoto flow still lacks a\ncomplete understanding, although there were some recent progress on the\ncomplete synchronization of the Kuramoto model in a sufficiently large coupling\nstrength regime. In this paper, we present sufficient frameworks for partial\nphase-locking of a majority ensemble and the complete phase-locking in terms of\nthe initial phase configuration, coupling strength and natural frequencies. As\na by-product of our analysis, we obtain nontrivial upper bounds for the\npathwise critical coupling strength in terms of the diameter of natural\nfrequencies, initial Kuramoto order parameter and the system size $N$. We also\nshow that phase-locked states whose order parameters are less than\n$N^{-\\frac{1}{2}}$ are linearly unstable.\n", " The tacnode Riemann-Hilbert problem is a 4 x 4 matrix valued RH problem that\nappears in the description of the local behavior of two touching groups of\nnon-intersecting Brownian motions. The same RH problem was also found by Duits\nand Geudens to describe a new critical regime in the two-matrix model.\n Delvaux gave integral representations for some of the entries of the 4 x 4\nmatrix. We complement this work by presenting integral representations for all\nof the entries. As a consequence we give an explicit formula for the\nDuits-Geudens critical kernel.\n", " We show that the higher derivatives of the Riemann zeta function may be\nexpressed in terms of integrals involving the digamma function. Related\nintegrals for the Stieltjes constants are also shown. We also present a formula\nfor the derivatives of the Riemann zeta function entirely in terms of the\nLehmer constants.\n", " Comprehensive and predictive simulation of coupled reaction networks has long\nbeen a goal of biology and other fields. Currently, metabolic network models\nthat utilize enzyme mass action kinetics have predictive power but are limited\nin scope and application by the fact that the determination of enzyme rate\nconstants is laborious and low throughput. We present a statistical\nthermodynamic formulation of the law of mass action for coupled reactions at\nboth steady states and non-stationary states. The formulation is based on a\nfluctuation theorem for coupled reactions and uses chemical potentials instead\nof rate constants. When used to model deterministic systems, the theory\ncorresponds to a rescaling of the time dependent reactions in such a way that\nsteady states can be reached on the same time scale but with significantly\nfewer computational steps. The significance for applications in systems biology\nis discussed.\n", " Zero temperature spectra of mesons and glueballs are analyzed in a class of\nholographic bottom-up models for QCD in the Veneziano limit, N_c -> infinity,\nN_f -> infinity, with x = N_f/N_c fixed (V-QCD). The backreaction of flavor on\ncolor is fully included. It is found that spectra are discrete and gapped\n(modulo the pions) in the QCD regime, for x below the critical value x_c where\nthe conformal transition takes place. The masses uniformly converge to zero in\nthe walking region x -> x_c^- due to Miransky scaling. All the ratios of masses\nasymptote to non-zero constants as x -> x_c^- and therefore there is no\n\"dilaton\" in the spectrum. The S-parameter is computed and found to be of O(1)\nin units of N_f N_c in the walking regime, while it is always an increasing\nfunction of x. This indicates the presence of a subtle discontinuity of\ncorrelation functions across the conformal transition at x = x_c.\n", " We study the $k$-th nearest neighbor distance function from a finite\npoint-set in $\\mathbb{R}^d$. We provide a Morse theoretic framework to analyze\nthe sub-level set topology. In particular, we present a simple\ncombinatorial-geometric characterization for critical points and their indices,\nalong with detailed information about the possible changes in homology at the\ncritical levels. We conclude by computing the expected number of critical\npoints for a homogeneous Poisson process. Our results deliver significant\ninsights and tools for the analysis of persistent homology in order-$k$\nDelaunay mosaics, and random $k$-fold coverage.\n", " In the paper, the principal aspects of the mathematical theory of equilibrium\nthermodynamics are distinguished. It is proved that the points of degeneration\nof a Bose gas of fractal dimension in the momentum space coincide with critical\npoints or real gases, whereas the jumps of critical indices and the Maxwell\nrule are related to the tunnel generalization of thermodynamics. Semiclassical\nmethods are considered for the tunnel generalization of thermodynamics and also\nfor the second and ultrasecond quantization (operators of creation and\nannihilation of pairs). To every pure gas there corresponds a new critical\npoint of the limit negative pressure below which the liquid passes to a\ndispersed state (a foam). Relations for critical points of a homogeneous\nmixture of pure gases are given in dependence on the concentration of gases.\n", " The problem of leaders location plays an important role in the\ncontrollability of undirected graphs.The concept of minimal perfect critical\nvertex set is introduced by drawing support from the eigenvector of Laplace\nmatrix. Using the notion of minimal perfect critical vertex set, the problem of\nfinding the minimum number of controllable leader vertices is transformed into\nthe problem of finding all minimal perfect critical vertex sets. Some necessary\nand sufficient conditions for special minimal perfect critical vertex sets are\nprovided, such as minimal perfect critical 2 vertex set, and minimal perfect\ncritical vertex set of path or path related graphs. And further, the leaders\nlocation problem for path graphs is solved completely by the algorithm provided\nin this paper. An interesting result that there never exist a minimal perfect\ncritical 3 vertex set is proved, too.\n", " A new proof is given for the fact that centered gaussian functions saturate\nthe Euclidean forward-reverse Brascamp-Lieb inequalities, extending the\nBrascamp-Lieb and Barthe theorems. A duality principle for best constants is\nalso developed, which generalizes the fact that the best constants in the\nBrascamp-Lieb and Barthe inequalities are equal. Finally, as the title hints,\nthe main results concerning finiteness, structure and gaussian-extremizability\nfor the Brascamp-Lieb inequality due to Bennett, Carbery, Christ and Tao are\ngeneralized to the setting of the forward-reverse Brascamp-Lieb inequality.\n", " A new formulation of perturbation theory for a description of the Dirac and\nscalar fields (the Yukawa model) is suggested. As the main approximation the\nself-consistent field model is chosen, which allows in a certain degree to\naccount for the effects caused by the interaction of fields. Such choice of the\nmain approximation leads to a normally ordered form of the interaction\nHamiltonian. Generation of the fermion mass due to the interaction with\nexchange of the scalar boson is investigated. It is demonstrated that, for zero\nbare mass, the fermion can acquire mass only if the coupling constant exceeds\nthe critical value determined by the boson mass. In this connection, the\nproblem of the neutrino mass is discussed.\n", " In this article, we first improve the scalar maximal theorem for the Dunkl\nmaximal operator by giving some precisions on the behavior of the constants of\nthis theorem for a general reflection group. Next we complete the vector-valued\ntheorem for the Dunkl-type Fefferman-Stein operator in the case $\\mathbb Z_2^d$\nby establishing a result of exponential integrability corresponding to the case\n$p=+\\infty$.\n", " The paper discusses sharp sufficient conditions for interpolation and\nsampling for functions of n variables with convex spectrum. When n=1, the\nclassical theorems of Ingham and Beurling state that the critical values in the\nestimates from above (from below) for the distances between interpolation\n(sampling) nodes are the same. This is no longer true for n>1. While the\ncritical value for sampling sets remains constant, the one for interpolation\ngrows linearly with the dimension.\n", " $J/\\Psi$-nuclear bound state energies are calculated for a range of nuclei by\nsolving the Proca (Klein-Gordon) equation. Critical input for the calculations,\nnamely the medium-modified $D$ and $D^*$ meson masses, as well as the nucleon\ndensity distributions in nuclei, are obtained from the quark-meson coupling\nmodel. The attractive potential originates from the $D$ and $D^*$ meson loops\nin the $J/\\Psi$ self-energy in nuclear medium. It appears that $J/\\Psi$-nuclear\nbound states should produce a clear experimental signature provided that the\n$J/\\Psi$ meson is produced in recoilless kinematics.\n", " We present a critical analysis of Pade-based methods for the unitarization of\nlow energy amplitudes. We show that the use of certain Pade Approximants to\ndescribe the resonance region may lead to inaccurate determinations. In\nparticular, we find that in the Linear Sigma Model the unitarization of the low\nenergy amplitude through the inverse amplitude method produces essentially\nincorrect results for the mass and width of the sigma. Alternative sequences of\nPades are studied and we find that the diagonal sequences (i.e., [N/N]) have\nmuch better convergence properties.\n", " In this work we consider the Kantowski-Sachs (KS) universe in the framework\nof varying speed of light theory. We present the general solutions of the\ngravitational field equations with variable speed of light $c(t)$,\ngravitational coupling parameter $G(t)$ and the decaying vacuum energy\n$\\Lambda(t)$ for the KS model. In the limiting case for the equation of state\n(EOS) parameter $\\gamma=2$ (stiff fluid with $p=\\rho c^2$) and $\\gamma =1$\n(dust with $p=0$), exact solutions of the field equations are obtained. The\nnumerical solutions are also presented for both the cases. Moreover, it is\nshown that in the limiting case of large time, the mean anisotropy parameter\ntends to zero for $\\gamma=2$ and $\\gamma=1$. Thus the time variation of the\nfundamental constants provides an effective mechanism for the isotropization of\nthe KS universe.\n", " The ring of symmetric functions $\\Lambda$, with natural basis given by the\nSchur functions, arise in many different areas of mathematics. For example, as\nthe cohomology ring of the grassmanian, and as the representation ring of the\nsymmetric group. One may define a coproduct on $\\Lambda$ by the plethystic\naddition on alphabets. In this way the ring of symmetric functions becomes a\nHopf algebra. The Littlewood--Richardson numbers may be viewed as the structure\nconstants for the co-product in the Schur basis. In the first part of this\nthesis we show that by using a generalization of the classical umbral calculus\nof Gian-Carlo Rota, one may deform the basis of Schur functions to find many\nother bases for which the Littlewood--Richardson numbers as coproduct structure\nconstants. The Macdonald polynomials are a somewhat mysterious qt-deformation\nof the Schur functions. The second part of this thesis contains a proof a\ngenerating function identity for the Macdonald polynomials which was originally\nconjectured by Kawanaka.\n", " There must exist a reformulation of quantum field theory, even at low\nenergies, which does not depend on classical time. The octonionic theory\nproposes such a reformulation, leading to a pre-quantum pre-spacetime theory.\nThe ingredients for constructing such a theory, which is also a unification of\nthe standard model with gravitation, are : (i) the pre-quantum theory of trace\ndynamics: a matrix-valued Lagrangian dynamics, (ii) the spectral action\nprinciple of non-commutative geometry, (iii) the number system known as the\noctonions, for constructing a non-commutative manifold and for defining\nelementary particles via Clifford algebras, (iv) a Lagrangian with $E_8 \\times\nE_8$ symmetry. The split bioctonions define a sixteen dimensional space (with\nleft-right symmetry) whose geometry (evolving in Connes time) relates to the\nfour known fundamental forces, while predicting two new forces, $SU(3)_{grav}$\nand $U(1)_{grav}$. This latter interaction is possibly the theoretical origin\nof MOND. Coupling constants of the standard model result from left-right\nsymmetry breaking, and their values are theoretically determined by the\ncharacteristic equation of the exceptional Jordan algebra of the octonions. The\nquantum-to-classical transition, precipitated by the entanglement of a critical\nnumber of fermions, is responsible for the emergence of classical spacetime,\nand also for the familiar formulation of quantum theory on a spacetime\nbackground.\n", " The critical density for kaon condensation in ``nuclear star\" matter is\ncomputed up to two-loop order {\\it in medium} (corresponding to\nnext-to-next-to-leading order in chiral perturbation theory in free space) with\na heavy-baryon effective chiral Lagrangian whose parameters are determined from\n$KN$ scattering and kaonic atom data. To the order considered, the kaon\nself-energy has highly non-linear density dependence in dense matter. We find\nthat the four-Fermi interaction terms in the chiral Lagrangian play an\nimportant role in triggering condensation, predicting for ``natural\" values of\nthe four-Fermi interactions a rather low critical density, $\\rho_c < 4 \\rho_0$.\n", " We consider two Lax systems for the homogeneous Painlev\\'{e} II equation: one\nof size $2\\times 2$ studied by Flaschka and Newell in the early 1980's, and one\nof size $4\\times 4$ introduced by Delvaux-Kuijlaars-Zhang and Duits-Geudens in\nthe early 2010's. We prove that solutions to the $4\\times 4$ system can be\nderived from those to the $2\\times 2$ system via an integral transform, and\nconsequently relate the Stokes multipliers for the two systems. As corollaries\nwe are able to express two kernels for determinantal processes as contour\nintegrals involving the Flaschka-Newell Lax system: the tacnode kernel arising\nin models of nonintersecting paths, and a critical kernel arising in a\ntwo-matrix model.\n", " Piecewise-linear (PL) Morse theory and discrete Morse theory are used in\nshape analysis tasks to investigate the topological features of discretized\nspaces. In spite of their common origin in smooth Morse theory, various notions\nof critical points have been given in the literature for the discrete setting,\nmaking a clear understanding of the relationships occurring between them not\nobvious. This paper aims at providing equivalence results about critical points\nof the two discretized Morse theories. First of all, we prove the equivalence\nof the existing notions of PL critical points. Next, under an optimality\ncondition called relative perfectness, we show a dimension agnostic\ncorrespondence between the set of PL critical points and that of discrete\ncritical simplices of the combinatorial approach. Finally, we show how a\nrelatively perfect discrete gradient vector field can be algorithmically built\nup to dimension 3. This way, we guarantee a formal and operative connection\nbetween critical sets in the PL and discrete theories.\n", " Three-dimensional Ising model in zero external field is exactly solved by\noperator algebras, similar to the Onsager's approach in two dimensions. The\npartition function of the simple cubic crystal imposed by the periodic boundary\ncondition along two directions and the screw boundary condition along the third\ndirection is calculated rigorously. In the thermodynamic limit an integral\nreplaces a sum in the formula of the partition function. The critical\ntemperatures, at which order-disorder transitions in the infinite crystal occur\nalong three axis directions, are determined. The analytical expressions for the\ninternal energy and the specific heat are also presented.\n", " The Light-front quark model (LFQM) has been applied to calculate the\ntransition matrix elements of heavy hadron decays. However, it is noted that\nusing the traditional wave functions of the LFQM given in literature, the\ntheoretically determined decay constants of the $\\Upsilon(nS)$ obviously\ncontradict to the data. It implies that the wave functions must be modified.\nKeeping the orthogonality among the $nS$ states and fitting their decay\nconstants we obtain a series of the wave functions for $\\Upsilon(nS)$. Based on\nthese wave functions and by analogy to the hydrogen atom, we suggest a modified\nanalytical form for the $\\Upsilon(nS)$ wave functions. By use of the modified\nwave functions, the obtained decay constants are close to the experimental\ndata. Then we calculate the rates of radiative decays of $\\Upsilon(nS)\\to\n\\eta_b+\\gamma$. Our predictions are consistent with the experimental data on\ndecays $\\Upsilon(3S)\\to \\eta_b+\\gamma$ within the theoretical and experimental\nerrors.\n", " We study an action of the skew divided difference operators on the Schubert\npolynomials and give an explicit formula for structural constants for the\nSchubert polynomials in terms of certain weighted paths in the Bruhat order on\nthe symmetric group. We also prove that, under certain assumptions, the skew\ndivided difference operators transform the Schubert polynomials into\npolynomials with positive integer coefficients.\n", " We update the constraints on the time variation of the fine structure\nconstant $\\alpha$ and the electron mass $m_e$, using the latest CMB data,\nincluding the 7-yr release of WMAP. We made statistical analyses of the\nvariation of each one of the constants and of their joint variation, together\nwith the basic set of cosmological parameters. We used a modified version of\nCAMB and COSMOMC to account for these possible variations. We present bounds on\nthe variation of the constants for different data sets, and show how results\ndepend on them. When using the latest CMB data plus the power spectrum from\nSloan Digital Sky Survey LRG, we find that $\\alpha / \\alpha_0=0.986 \\pm 0.007$\nat 1-$\\sigma$ level, when the 6 basic cosmological parameters were fitted, and\nonly variation in $\\alpha$ was allowed. The constraints in the case of\nvariation in both constants are $ \\alpha / \\alpha_0= 0.986 \\pm 0.009$ and $m_e\n/ m_{e0} = 0.999 \\pm 0.035$. In the case of only variation in $m_e$, the bound\nis $m_e /m_{e0}=0.964 \\pm 0.025$.\n", " Superflow in a phenomenological tight-binding model for the superconducting\nstate of some High-temperature superconductors is discussed thoroughly. The\nformalism used is explicitly gauge-invariant and currents are computed exactly\nwithin BCS theory, going therefore beyond linear response theory. The\ndependence of gap functions, current density, critical currents, and free\nenergy as a function of superfluid velocity for different angles and doping\nconcentrations is investigated. Different sources of anisotropy, like the\ndispersion relation of the model, the internal symmetry of the order parameter,\nand orthorhombic distortions of the lattice are also studied.\n", " Computational thinking is a new problem soling method named for its extensive\nuse of computer science techniques. It synthesizes critical thinking and\nexisting knowledge and applies them in solving complex technological problems.\nThe term was coined by J. Wing, but the relationship between computational and\ncritical thinking, the two modes of thiking in solving problems, has not been\nyet learly established. This paper aims at shedding some light into this\nrelationship. We also present two classroom experiments performed recently at\nthe Graduate Technological Educational Institute of Patras in Greece. The\nresults of these experiments give a strong indication that the use of computers\nas a tool for problem solving enchances the students' abilities in solving real\nworld problems involving mathematical modelling. This is also crossed by\nearlier findings of other researchers for the problem solving process in\ngeneral (not only for mathematical problems).\n", " We prove a one-parameter family of sharp integral inequalities for functions\non the $n$-dimensional unit ball. The inequalities are conformally invariant,\nand the sharp constants are attained for functions that are equivalent to a\nconstant function under conformal transformations. As a limiting case, we\nobtain an inequality that generalizes Carleman's inequality for harmonic\nfunctions in the plane to poly-harmonic functions in higher dimensions.\n", " A practical high-accuracy relativistic method of atomic structure\ncalculations for univalent atoms is presented. The method is rooted in the\ncoupled-cluster formalism and includes non-perturbative treatment of single and\ndouble excitations from the core and single, double and triple excitations\ninvolving valence electron. Triple excitations of core electrons are included\nin the fourth-order of many-body perturbation theory. In addition,\ncontributions from the disconnected excitations are incorporated. Evaluation of\nmatrix elements includes all-order dressing of lines and vertices of the\ndiagrams. The resulting formalism for matrix elements is complete through the\nfourth order and sums certain chains of diagrams to all orders. With the\ndeveloped method we compute removal energies, magnetic-dipole\nhyperfine-structure constants A and electric-dipole amplitudes. We find that\nthe removal energies are reproduced within 0.01-0.03% and the hyperfine\nconstants of the 3s_1/2 and 3p_1/2 states with a better than 0.1% accuracy. The\ncomputed dipole amplitudes for the principal 3s_1/2 - 3p_1/2;3/2 transitions\nare in an agreement with 0.05%-accurate experimental data.\n", " We compute the best constant in functional integral inequality called the\nHardy-Leray inequalities for solenoidal vector fields on $\\mathbb{R}^N$. This\ngives a solenoidal improvement of the inequalities whose best constants are\nknown for unconstrained fields, and develops of the former work by\nCostin-Maz'ya who found the best constant in the Hardy-Leray inequality for\naxisymmetric solenoidal fields. We derive the same best constant without any\nsymmetry assumption, whose expression can be simplified in relation to the\nweight exponent. Moreover, it turns out that the best value cannot be attained\nin the space of functions satisfying the finiteness of the integrals in the\ninequality.\n", " More than a century has passed since van't Hoff and Arrhenius formulated\ntheir celebrated rate theories, but there are still elusive aspects in the\ntemperature-dependent phase transition kinetics of molecular systems. Here I\npresent a theory based on microcanonical thermostatistics that establishes a\nsimple and direct temperature dependence for all rate constants, including the\nforward and the reverse rate constants, the equilibrium constant, and the\nnucleation rate. By considering a generic model that mimic the microcanonical\ntemperature of molecular systems in a region close to a first-order phase\ntransition, I obtain shape-free relations between kinetics and thermodynamics\nphysical quantities which are validated through stochastic simulations.\nAdditionally, the rate theory is applied to results obtained from protein\nfolding and ice nucleation experiments, demonstrating that the expressions\nderived here can be used to describe the experimental data of a wide range of\nmolecular systems.\n", " An equivalent, but variant form of the Riemann functional equation is\nexplored, and several discoveries are made. Properties of the Riemann zeta\nfunction $\\zeta(s)$ from which a necessary and sufficient condition for the\nexistence of zeros in the critical strip are deduced. This in turn, by an\nindirect route, eventually produces a simple, solvable, differential equation\nfor $arg(\\zeta(s))$ on the critical line $s=1/2+i\\rho$, the consequences of\nwhich are explored, and the \"LogZeta\" function is introduced. A singular linear\ntransform between the real and imaginary components of $\\zeta$ and\n$\\zeta^\\prime$ on the critical line is derived, and an implicit relationship\nfor locating a zero ($\\rho=\\rho_0$) on the critical line is found between the\narguments of $\\zeta(1/2+i\\rho)$ and $\\zeta^{\\prime}(1/2+i\\rho)$. Notably, the\nVolchkov criterion, a Riemann Hypothesis (RH) equivalent is analytically\nevaluated and verified to be half equivalent to RH, but RH is not proven.\nNumerical results are presented, some of which lead to the identification of\n{\\it anomalous zeros}, whose existence in turn suggests that well-established,\ntraditional derivations such as the Volchkov criterion and counting theorems\nrequire re-examination. It is proven that the derivative\n$\\zeta^{\\prime}(1/2+i\\rho)$ will never vanish on the perforated critical line\n($\\rho\\neq\\rho_0$). Traditional asymptotic and counting results are obtained in\nan untraditional manner, yielding insight into the nature of $\\zeta(1/2+i\\rho)$\nas well as very accurate asymptotic estimates for distribution bounds and the\ndensity of zeros on the critical line.\n", " We work in a discrete model of the nonlinear Fourier transform (following the\nterminology of Tao and Thiele), which appears in the study of orthogonal\npolynomials on the unit circle. The corresponding nonlinear variant of the\nHausdorff-Young inequality can be deduced by adapting the ideas of Christ and\nKiselev to the present discrete setting. However, the behavior of sharp\nconstants remains largely unresolved. In this short note we give two results on\nthese constants, after restricting our attention to either sufficiently small\nsequences or to sequences that are far from being the extremizers of the linear\nHausdorff-Young inequality.\n", " The discrete orthogonality relations for the multi-indexed orthogonal\npolynomials in discrete quantum mechanics with pure imaginary shifts are\ninvestigated. We show that the discrete orthogonality relations hold for the\ncase-(1) multi-indexed orthogonal polynomials of continuous Hahn, Wilson and\nAskey-Wilson types, and conjecture their normalization constants.\n", " For polynomial ODE models, we introduce and discuss the concepts of exact and\napproximate conservation laws, which are the first integrals of the full and\ntruncated sets of ODEs. For fast-slow systems, truncated ODEs describe the fast\ndynamics. We define compatibility classes as subsets of the state space,\nobtained by equating the conservation laws to constants. A set of conservation\nlaws is complete when the corresponding compatibility classes contain a finite\nnumber of steady states. Complete sets of conservation laws can be used for\nmodel order reduction and for studying the multistationarity of the model. We\nprovide algorithmic methods for computing linear, monomial, and polynomial\nconservation laws of polynomial ODE models and for testing their completeness.\nThe resulting conservation laws and their completeness are either independent\nor dependent on the parameters. In the latter case, we provide parametric case\ndistinctions. In particular, we propose a new method to compute polynomial\nconservation laws by comprehensive Gr\\\"obner systems and syzygies.\n Keywords: First integrals, chemical reaction networks, polynomial\nconservation laws, syzygies, comprehensive Gr\\\"obner systems.\n", " We study the gauge dependence of the neutrino mass renormalization in a two\nHiggs doublet model, that is extended with one singlet seesaw neutrino. This\nmodel gives only one light neutrino a mass at tree level, while the second\nlight mass is generated at loop level via the interaction with the second Higgs\ndoublet. At one loop level, one neutrino stays massless. We use multiplicative\nrenormalization constants to define counterterms. The renormalized mass\nparameters are defined as the complex poles of the propagators, using the\ncomplex mass scheme for mass renormalization. With this setup, we analytically\nget the expressions for the neutrino mass counterterms and isolate the gauge\ndependent part. We show, how relating this gauge dependent part with the\ntadpole renormalization leads to gauge independent counterterm definitions,\nhence gauge independent bare masses for neutrinos.\n", " In this work interacting Umami Chaplygin gas has been studied in flat FRW\nmodel of universe in context of it's thermodynamic and dynamical behaviour. In\nparticular, considering Umami fluid as dark energy interacting with dark\nmatter, irreversible thermodynamics has been studied both for apparent and\nevent horizon as bounding horizon in two separate cases. Also the model has\nbeen investigated in purview of dynamical systems analysis by converting the\ncosmological evolution equations to an autonomous system of ordinary\ndifferential equations. With some restrictions on model parameter $\\omega$ and\ncoupling parameter $\\lambda$, some cosmologically interesting critical points\ndescribing late time accelerated evolution of the universe attracted by\ncosmological constant and accelerated scaling attractor in quintessence era\nhave been found to alleviate coincidence problem.\n", " In this work we show that for every $d < \\infty$ and the Ising model defined\non $G(n,d/n)$, there exists a $\\beta_d > 0$, such that for all $\\beta <\n\\beta_d$ with probability going to 1 as $n \\to \\infty$, the mixing time of the\ndynamics on $G(n,d/n)$ is polynomial in $n$. Our results are the first\npolynomial time mixing results proven for a natural model on $G(n,d/n)$ for $d\n> 1$ where the parameters of the model do not depend on $n$. They also provide\na rare example where one can prove a polynomial time mixing of Gibbs sampler in\na situation where the actual mixing time is slower than $n \\polylog(n)$. Our\nproof exploits in novel ways the local treelike structure of Erd\\H{o}s-R\\'enyi\nrandom graphs, comparison and block dynamics arguments and a recent result of\nWeitz.\n Our results extend to much more general families of graphs which are sparse\nin some average sense and to much more general interactions. In particular,\nthey apply to any graph for which every vertex $v$ of the graph has a\nneighborhood $N(v)$ of radius $O(\\log n)$ in which the induced sub-graph is a\ntree union at most $O(\\log n)$ edges and where for each simple path in $N(v)$\nthe sum of the vertex degrees along the path is $O(\\log n)$. Moreover, our\nresult apply also in the case of arbitrary external fields and provide the\nfirst FPRAS for sampling the Ising distribution in this case. We finally\npresent a non Markov Chain algorithm for sampling the distribution which is\neffective for a wider range of parameters. In particular, for $G(n,d/n)$ it\napplies for all external fields and $\\beta < \\beta_d$, where $d \\tanh(\\beta_d)\n= 1$ is the critical point for decay of correlation for the Ising model on\n$G(n,d/n)$.\n", " In this work, we present a phenomenological study of the complete analytical\nsolution to the bound eigenstates and eigenvalues of the Yukawa potential\nobtained previously using the hidden supersymmetry of the system and a\nsystematic expansion of the Yukawa potential in terms of $\\delta=a_{0}/D$,\nwhere $a_{0}$ is the Bohr radius and $D$ is the screening length. The\neigenvalues, $\\epsilon_{nl}(\\delta)$, are given in the form of Taylor series in\n$\\delta$ which can be systematically calculated to the desired order\n$\\delta^{k}$. Coulomb $l$-degeneracy is broken by the screening effects and,\nfor a given $n$, $\\epsilon_{nl}(\\delta)$ is larger for higher values of $l$\nwhich causes the crossing of levels for $n\\ge4$. The convergence radius of the\nTaylor series can be enlarged up to the critical values using the Pad\\'e\napproximants technique which allows us to calculate the eigenvalues with high\nprecision in the whole rage of values of $\\delta$ where bound states exist, and\nto reach a precise determination of the critical screening lengths,\n$\\delta_{nl}$. Eigenstates have a form similar to the solutions of the Coulomb\npotential, with the associated Laguerre polynomials replaced by new polynomials\nof order $\\delta^{k}$ with $r$-dependent coefficients which, in turn, are\npolynomials in $r$. In general we find sizable deviations from the Coulomb\nradial probabilities only for screening lengths close to their critical values.\nWe use these solutions to find the squared absolute value at the origin of the\nwave function for $l=0$, and their derivatives for $l=1$, for the lowest\nstates, as functions of $\\delta$, which enter the phenomenology of dark matter\nbound states in dark gauge theories with a light dark mediator.\n", " In this work we prove some Hardy-Poincar\\'{e} inequalities with quadratic\nsingular potentials localized on the boundary of a smooth domain. Then, we\nconsider conical domains with vertex on the singularity and we show upper and\nlower bounds for the corresponding optimal constants in the Hardy inequality.\nIn particular, we prove the asymptotic behavior of the optimal constant when\nthe slot of the cone tends to zero.\n", " We extend earlier work on fully symmetric polynomials for three-boson wave\nfunctions to arbitrarily many bosons and apply these to a light-front analysis\nof the low-mass eigenstates of $\\phi^4$ theory in 1+1 dimensions. The\nbasis-function approach allows the resolution in each Fock sector to be\nindependently optimized, which can be more efficient than the preset discrete\nFock states in DLCQ. We obtain an estimate of the critical coupling for\nsymmetry breaking in the positive mass-squared case.\n", " In this paper we lay special stress on analyzing the topological properties\nof the lattice systems and try to ovoid the conventional ways to calculate the\ncritical points. Only those clusters with finite sizes can execute the self\nsimilar transformations of infinite hierarchies. Each ordered cluster has\nfractal dimension, their minimum relates to the edge fixed point, which accords\nwith the transformations fixed point relating to a critical point. There are\ntwo classes of systems and clusters by their connectivity. Using mathematic\nmapping method we set up cluster-spin Gaussian model solved accurately. There\nare single state and -fold coupling state in a reducible cluster, each of which\ncorresponds to a subsystem described by a Gaussian model. By the minimal\nfractal dimension a final expression of the critical points is obtained. The\ncritical points of three lattice systems are calculated, our method makes the\ncalculations very simplified and highly accurate. A possibility of existence of\ndifferent clusters at the critical temperature is discussed.\n", " We demonstrate existence and uniqueness of Picard--Vessiot extensions\nsatisfying prescribed properties, for systems of linear differential equations\nover a field satisfying the same properties, under some closure assumptions on\nthe field of constants. An example includes the case of a equations over a real\nfield, with a real-closed field of constants. The result is obtained through a\nmodel theoretic approach.\n", " We study aperiodic balanced sequences over finite alphabets. A sequence vv of\nthis type is fully characterised by a Sturmian sequence u and two constant gap\nsequences y and y'. We show that the language of v is eventually dendric and we\nfocus on return words to its factors. We develop a method for computing the\ncritical exponent and asymptotic critical exponent of balanced sequences,\nprovided the associated Sturmian sequence u has a quadratic slope. The method\nis based on looking for the shortest return words to bispecial factors in v. We\nillustrate our method on several examples; in particular we confirm a\nconjecture of Rampersad, Shallit and Vandomme that two specific sequences have\nthe least critical exponent among all balanced sequences over 9-letter (resp.,\n$0-letter) alphabets.\n", " Define a boundary point of a graph which is embedded in the Euclidean plane a\nvertex which is incident to only one edge. In this paper we consider graphs\nwhich are embedded in the Euclidean plane with a finite number of boundary\npoints. The simple geometric condition we impose on them is that the sum of\nunit vectors of edges extending from each non-boundary vertex will be equal to\nzero. We call such a graph a critical graph and ask to maximise the number of\nvertices of critical graphs with a given size of boundary. The main results of\nthis paper give a sharp upper bound for the maximum number of vertices of\nplanar critical graphs, where the degree of each non-boundary vertex is 3 or 4.\n", " Given a (directed) graph G=(V,A), a subset X of V is an interval of G\nprovided that for any a, b\\in X and x\\in V-X, (a,x)\\in A if and only if\n(b,x)\\in A and (x,a)\\in A if and only if (x,b)\\in A. For example, \\emptyset,\n\\{x\\} (x \\in V) and V are intervals of G, called trivial intervals. A graph,\nall the intervals of which are trivial, is indecomposable; otherwise, it is\ndecomposable. A vertex x of an indecomposable graph is critical if G-x is\ndecomposable. In 1993, J.H. Schmerl and W.T. Trotter characterized the\nindecomposable graphs, all the vertices of which are critical, called critical\ngraphs. In this article, we characterize the indecomposable graphs which admit\na single non critical vertex, that we call (-1)-critical graphs.} This gives an\nanswer to a question asked by Y. Boudabbous and P. Ille in a recent article\nstudying the critical vertices in an indecomposable graph.\n", " Motivated by automated junction recognition in tracking data, we study a\nproblem of placing a square or disc of fixed size in an arrangement of lines or\nline segments in the plane. We let distances among the intersection points of\nthe lines and line segments with the square or circle define a clustering, and\nstudy the complexity of \\emph{critical} placements for this clustering. Here\ncritical means that arbitrarily small movements of the placement change the\nclustering.\n A parameter $\\varepsilon$ defines the granularity of the clustering. Without\nany assumptions on $\\varepsilon$, the critical placements have a trivial\n$O(n^4)$ upper bound. When the square or circle has unit size and $0 <\n\\varepsilon < 1$ is given, we show a refined $O(n^2/\\varepsilon^2)$ bound,\nwhich is tight in the worst case.\n We use our combinatorial bounds to design efficient algorithms to compute\njunctions. As a proof of concept for our algorithms we have a prototype\nimplementation that showcases their application in a basic visualization of a\nset of $n$ trajectories and their $k$ most important junctions.\n", " Recent years have witnessed a large amount of decentralized data in multiple\n(edge) devices of end-users, while the aggregation of the decentralized data\nremains difficult for machine learning jobs due to laws or regulations.\nFederated Learning (FL) emerges as an effective approach to handling\ndecentralized data without sharing the sensitive raw data, while\ncollaboratively training global machine learning models. The servers in FL need\nto select (and schedule) devices during the training process. However, the\nscheduling of devices for multiple jobs with FL remains a critical and open\nproblem. In this paper, we propose a novel multi-job FL framework to enable the\nparallel training process of multiple jobs. The framework consists of a system\nmodel and two scheduling methods. In the system model, we propose a parallel\ntraining process of multiple jobs, and construct a cost model based on the\ntraining time and the data fairness of various devices during the training\nprocess of diverse jobs. We propose a reinforcement learning-based method and a\nBayesian optimization-based method to schedule devices for multiple jobs while\nminimizing the cost. We conduct extensive experimentation with multiple jobs\nand datasets. The experimental results show that our proposed approaches\nsignificantly outperform baseline approaches in terms of training time (up to\n8.67 times faster) and accuracy (up to 44.6% higher).\n", " Capabilities for collecting and storing data on mobile objects have increased\ndramatically over the past few decades. A persistent difficulty is summarizing\nlarge collections of mobile objects. This paper develops methods for extracting\nand analyzing hotspots or locations with relatively high levels of mobility\nactivity. We use kernel density estimation (KDE) to convert a large collection\nof mobile objects into a smooth, continuous surface. We then develop a\ntopological algorithm to extract critical geometric features of the surface;\nthese include critical points (peaks, pits and passes) and critical lines\n(ridgelines and course-lines). We connect the peaks and corresponding\nridgelines to produce a surface network that summarizes the topological\nstructure of the surface. We apply graph theoretic indices to analytically\ncharacterize the surface and its changes over time. To illustrate our approach,\nwe apply the techniques to taxi cab data collected in Shanghai, China. We find\nincreases in the complexity of the hotspot spatial distribution during normal\nactivity hours in the late morning, afternoon and evening and a spike in the\nconnectivity of the hotspot spatial distribution in the morning as taxis\nconcentrate on servicing travel to work. These results match with scientific\nand anecdotal knowledge about human activity patterns in the study area.\n", " The dispersive analysis of the decay $\\eta\\to3\\pi$ is reviewed and thoroughly\nupdated with the aim of determining the quark mass ratio\n~$Q^2=(m_s^2-m_{ud}^2)/(m_d^2-m_u^2)$. With the number of subtractions we are\nusing, the effects generated by the final state interaction are dominated by\nlow energy $\\pi\\pi$ scattering. Since the corresponding phase shifts are now\naccurately known, causality and unitarity determine the decay amplitude within\nsmall uncertainties -- except for the values of the subtraction constants. Our\ndetermination of these constants relies on the Dalitz plot distribution of the\ncharged channel, which is now measured with good accuracy. The theoretical\nconstraints that follow from the fact that the particles involved in the\ntransition represent Nambu-Goldstone bosons of a hidden approximate symmetry\nplay an equally important role. The ensuing predictions for the Dalitz plot\ndistribution of the neutral channel and for the branching ratio\n$\\Gamma_{\\eta\\to3\\pi^0}/ \\Gamma_{\\eta\\to\\pi^+\\pi^-\\pi^0}$ are in very good\nagreement with experiment. Relying on a known low-energy theorem that relates\nthe meson masses to the masses of the three lightest quarks, our analysis leads\nto $Q=22.1(7)$, where the error covers all of the uncertainties encountered in\nthe course of the calculation: experimental uncertainties in decay rates and\nDalitz plot distributions, noise in the input used for the phase shifts, as\nwell as theoretical uncertainties in the constraints imposed by chiral symmetry\nand in the evaluation of isospin breaking effects. Our result indicates that\nthe current algebra formulae for the meson masses only receive small\ncorrections from higher orders of the chiral expansion, but not all of the\nrecent lattice results are consistent with this conclusion.\n" ]

[ " We present the results of experimental and theoretical studies of fast\nanisotropic relaxation and rotational diffusion in the first electron excited\nstate of biological coenzyme NADH in water-ethanol solutions. The experiments\nhave been carried out by means of a novel polarization-modulation transient\nmethod and fluorescence polarization spectroscopy. For interpretation of the\nexperimental results a model of the anisotropic relaxation in terms of scalar\nand vector properties of transition dipole moments and based on the\nBorn-Oppenheimer approximation has been developed. The results obtained suggest\nthat the dynamics of anisotropic rovibronic relaxation in NADH under excitation\nwith 100~fs pump laser pulses can be characterised by a single vibration\nrelaxation time $\\tau_v$ laying in the range 2--15~ps and a single rotation\ndiffusion time $\\tau_r$ laying in the range 100--450~ps a subject of ethanol\nconcentration. The dependence of the times $\\tau_v$ and $\\tau_r$ on the\nsolution polarity (static permittivity) and viscosity has been determined and\nanalyzed. Limiting values of an important parameter $\\langle\nP_2(\\cos\\theta(t))\\rangle$ describing the rotation of the transition dipole\nmoment in the course of vibrational relaxation has been determined from\nexperiment as function of the ethanol concentration and analyzed.\n", " Fascinating developments in optical pulse engineering over the last 20 years\nlead to the generation of laser pulses as short as few femtosecond, providing a\nunique tool for high resolution time domain spectroscopy. However, a number of\nthe processes in nature evolve with characteristic times of the order of 1 fs\nor even shorter. Time domain studies of such processes require at first place\nsub-fs resolution, offered by pulse depicting attosecond localization. The\ngeneration, characterization and proof of principle applications of such pulses\nis the target of the attoscience.\n In the paper the thermal processes on the attosecond scale are described. The\nKlein-Gordon and Proca equations are developed. The relativistic effects in the\nheat transport on nanoscale are discussed. It is shown that the standard\nFourier equation can not be valid for the transport phenomena induced by\nattosecond laser pulses. The heat transport in nanoparticles and nanotubules is\ninvestigated.\n", " Protein interaction networks aim to summarize the complex interplay of\nproteins in an organism. Early studies suggested that the position of a protein\nin the network determines its evolutionary rate but there has been considerable\ndisagreement as to what extent other factors, such as protein abundance, modify\nthis reported dependence.\n We compare the genomes of Saccharomyces cerevisiae and Caenorhabditis elegans\nwith those of closely related species to elucidate the recent evolutionary\nhistory of their respective protein interaction networks. Interaction and\nexpression data are studied in the light of a detailed phylogenetic analysis.\nThe underlying network structure is incorporated explicitly into the\nstatistical analysis.\n The increased phylogenetic resolution, paired with high-quality interaction\ndata, allows us to resolve the way in which protein interaction network\nstructure and abundance of proteins affect the evolutionary rate. We find that\nexpression levels are better predictors of the evolutionary rate than a\nprotein's connectivity. Detailed analysis of the two organisms also shows that\nthe evolutionary rates of interacting proteins are not sufficiently similar to\nbe mutually predictive.\n It appears that meaningful inferences about the evolution of protein\ninteraction networks require comparative analysis of reasonably closely related\nspecies. The signature of protein evolution is shaped by a protein's abundance\nin the organism and its function and the biological process it is involved in.\nIts position in the interaction networks and its connectivity may modulate this\nbut they appear to have only minor influence on a protein's evolutionary rate.\n", " The olive oil sector produces a substantial impact in the Mediterranean's\neconomy and lifestyle. Many studies exist which try to optimize the different\nsteps in the olive oil's production process. One of the main challenges for\nolive oil producers is the ability to asses and control the quality during the\nproduction cycle. For this purpose, several parameters need to be determined,\nsuch as the acidity, the UV absorption or the ethyl esters content. To achieve\nthis, samples must be sent to an approved laboratory for chemical analysis.\nThis approach is expensive and cannot be performed very frequently, making\nquality control of olive oil a real challenge. This work explores a new\napproach based on fluorescence spectroscopy and artificial intelligence\n(namely, 1-D convolutional neural networks) to predict the five chemical\nquality indicators of olive oil (acidity, peroxide value, UV spectroscopic\nparameters $K_{270}$ and $K_{232}$, and ethyl esters) from simple fluorescence\nspectra. Fluorescence spectroscopy is a very attractive optical technique since\nit does not require sample preparation, is non destructive, and, as shown in\nthis work, can be easily implemented in small and cost-effective sensors. The\nresults indicate that the proposed approach gives exceptional results in the\nquality determination and would make the continuous quality control of olive\noil during and after the production process a reality. Additionally, this novel\nmethodology presents potential applications as a support for quality\nspecifications of olive oil, as defined by the European regulation.\n", " Photothermal-Induced Resonance (PTIR) is increasingly used in the measurement\nof infrared absorption spectra of sub-micrometer objects. The technique\nmeasures IR absorption spectra by relying on the photothermal effect induced by\na rapid pulse of light and the excitation of the resonance spectrum of an AFM\ncantilever in contact with the sample. In this work we assess the spatial\nresolution and depth response of PTIR in resonant mode while systematically\nvarying the pulsing frequency of the excitation laser and the cantilever\nresonance. The spatial resolution shows a shallow linear dependence on the\ninverse of the pulse frequency, which rules out tip size as a limiting factor\nfor resolution in the frequency range under investigation. Measured resolution\nvalues are also one order of magnitude lower than in the thermal diffusion\nlimit, excluding thermal wave propagation as a limiting factor. We also show\nthat the pulsing frequency of the laser and choice of cantilever resonance\naffect the intensity of the signal and the surface selectivity in PTIR images,\nwith higher frequencies providing increased surface selectivity. The results\nconfirm a difference in signal generation between resonant PTIR and other\nphotothermal techniques and indicate that photothermal induced heating and\nexpansion cannot fully account for observed intensity and resolution.\n", " Optical properties and morphology of laser generated Aluminum and Copper\nphthalocyanine nanoparticles (nAlPc and nCuPc) in water are experimentally\nstudied. Near infrared laser source of nanosecond pulse duration was used for\nfragmentation of Pc micro-powder suspended in H2O. Extinction spectra in the\nvisible and near IR range of NPs colloidal solutions in MQ water were acquired\nby means of optical spectroscopy. The optical density of both nCuPc and nAlPc\nincreases with laser fragmentation time. Transmission electron microscopy was\nused for characterization of nanoparticle morphology and size analysis. It is\nfound that nCuPc are made of short (100 nm) rectangular bars interconnected at\nvarious angles with other bars. Similar experiments were carried out for a\ncolloidal solution, which is a mixture of Au and AlPc nanoparticles. It turned\nout that Au NPs in presence of nAlPc form large agglomerates of Au.\n", " The measurement of the ultra high energy cosmic ray (UHECR) spectrum is\nstrongly affected by uncertainties on the reconstructed energy. The\ndetermination of the presence or absence of the GZK cutoff and its position in\nthe energy spectrum depends not only on high statistics but also on the shape\nof the energy error distribution. Here we determine the energy error\ndistribution for fluorescence telescopes, based on a Monte Carlo simulation.\nThe HiRes and Auger fluorescence telescopes are simulated in detail. We analyze\nthe UHECR spectrum convolved with this energy error distribution. We compare\nthis spectrum with one convolved with a lognormal error distribution as well as\nwith a Gaussian error distribution. We show that the energy error distribution\nfor fluorescence detectors can not be represented by these known distributions.\nWe conclude that the convolved energy spectrum will be smeared but not enough\nto affect the GZK cutoff detection. This conclusion stands for both HiRes and\nAuger fluorescence telescopes. This result differs from the effect of the\nenergy error distribution obtained with ground detectors and reinforces the\nimportance of the fluorescence energy measurement. We also investigate the\neffect of possible fluorescence yield measurement errors in the energy\nspectrum.\n", " The Casimir force is a universal interaction induced by electromagnetic\nquantum fluctuations between any types of objects. The expansion of the\ngraphene family by adding silicene, germanene, and stanene, 2D allotropes of\nSi, Ge, and Sn, lands itself as a platform to probe Dirac-like physics in\nhoneycomb staggered systems in such a ubiquitous interaction. We discover\nCasimir force phase transitions between these staggered 2D materials induced by\nthe complex interplay between Dirac physics, spin-orbit coupling, and\nexternally applied fields. In particular, we find that the interaction energy\nexperiences different power law distance decays, magnitudes, and dependences on\ncharacteristic physical constants. Furthermore, due to the topological\nproperties of these materials, repulsive and quantized Casimir interactions\nbecome possible.\n", " Artificial Spin Ice (ASI), consisting of a two dimensional array of nanoscale\nmagnetic elements, provides a fascinating opportunity to observe the physics of\nout of equilibrium systems. Initial studies concentrated on the static, frozen\nstate, whilst more recent studies have accessed the out-of-equilibrium dynamic,\nfluctuating state. This opens up exciting possibilities such as the observation\nof systems exploring their energy landscape through monopole quasiparticle\ncreation, potentially leading to ASI magnetricity, and to directly observe\nunconventional phase transitions. In this work we have measured and analysed\nthe magnetic relaxation of thermally active ASI systems by means of SQUID\nmagnetometry. We have investigated the effect of the interaction strength on\nthe magnetization dynamics at different temperatures in the range where the\nnanomagnets are thermally active and have observed that they follow an\nArrhenius-type N\\'eel-Brown behaviour. An unexpected negative correlation of\nthe average blocking temperature with the interaction strength is also\nobserved, which is supported by Monte Carlo simulations. The magnetization\nrelaxation measurements show faster relaxation for more strongly coupled\nnanoelements with similar dimensions. The analysis of the stretching exponents\nobtained from the measurements suggest 1-D chain-like magnetization dynamics.\nThis indicates that the nature of the interactions between nanoelements lowers\nthe dimensionality of the ASI from 2-D to 1-D. Finally, we present a way to\nquantify the effective interaction energy of a square ASI system, and compare\nit to the interaction energy calculated from a simple dipole model and also to\nthe magnetostatic energy computed with micromagnetic simulations.\n", " Fixing molecules in space is a crucial step for the imaging of molecular\nstructure and dynamics. Here, we demonstrate three-dimensional (3D) field-free\nalignment of the prototypical asymmetric top molecule indole using elliptically\npolarized, shaped, off-resonant laser pulses. A truncated laser pulse is\nproduced using a combination of extreme linear chirping and controlled phase\nand amplitude shaping using a spatial-light-modulator (SLM) based pulse shaper\nof a broadband laser pulse. The angular confinement is detected through\nvelocity-map imaging of H$^+$ and C$^{2+}$ fragments resulting from\nstrong-field ionization and Coulomb explosion of the aligned molecules by\nintense femtosecond laser pulses. The achieved three-dimensional alignment is\ncharacterized by comparing the result of ion-velocity-map measurements for\ndifferent alignment directions and for different times during and after the\nalignment laser pulse to accurate computational results. The achieved strong\nthree-dimensional field-free alignment of $\\langle \\cos^{2}\\delta \\rangle=0.89$\ndemonstrates the feasibility of both, strong three-dimensional alignment of\ngeneric complex molecules and its quantitative characterization.\n", " It is proved that when the retarded effect (or multiple moment effect) of\nradiation fields is taken into account, the high order stimulated radiation and\nstimulated absorption probabilities of light are not the same so that time\nreversal symmetry would be violated, though the total Hamiltonian of\nelectromagnetic interaction is still unchanged under time reversal. The reason\nto cause time reversal symmetry violation is that certain filial or partial\ntransitive processes of bounding state atoms are forbidden or can not be\nachieved actually due to the law of energy conservation, the asymmetric actions\nof effective transition operators before and after time reversal, as well as\nthe special states of atoms themselves. These restrictions would cause the\nsymmetry violation of time reversal of other filial or partial transition\nprocesses which can be actualized really. The symmetry violation is also\nrelative to the initial state asymmetries of bounding atoms before and after\ntime reversal. For the electromagnetic interaction between non-bounding state\natoms and radiation field, there is no this kind of symmetry violation of time\nreversal. In this way, the current formula of light stimulated radiation and\nabsorption parameters with time reversal symmetry should be revised. The\ninfluence of time reversal symmetry violation on the foundational theory of\nlaser is also discussed and the phenomena of non-population inversion and\nnon-radiation transition can also be explained well. In this way, a more\nreliable foundation can be established for the theories of laser and nonlinear\noptics in which non-equilibrium processes are involved.\n", " We theoretically investigate equilibrium behaviors and photothermal effects\nof a flexible plasmonic metamaterial composed of aramid nanofibers and gold\nnanoparticles. The fiber matrix is considered as an external field to\nreconfigure a nanoparticle assembly. We find that the heating process tunes\nparticle-particle and fiber-particle interactions, which alter adsorption of\nnanoparticles on fiber surfaces or clustering in pore spaces. Thus, it is\npossible to control the nanoparticle self-assembly by laser illumination. Gold\nnanoparticles strongly absorb radiations and efficiently dissipate absorbed\nenergy into heat. By solving the heat transfer equation associated with an\neffective medium approximation, we calculate the spatial temperature rise.\nRemarkably, our theoretical results quantitatively agree with prior\nexperiments. This indicates that we can ignore plasmonic coupling effects\ninduced by particle clustering. Effects of the laser spot size and intensity on\nthe photothermal heating are also discussed.\n", " Here we present ComPPI, a cellular compartment specific database of proteins\nand their interactions enabling an extensive, compartmentalized protein-protein\ninteraction network analysis (http://ComPPI.LinkGroup.hu). ComPPI enables the\nuser to filter biologically unlikely interactions, where the two interacting\nproteins have no common subcellular localizations and to predict novel\nproperties, such as compartment-specific biological functions. ComPPI is an\nintegrated database covering four species (S. cerevisiae, C. elegans, D.\nmelanogaster and H. sapiens). The compilation of nine protein-protein\ninteraction and eight subcellular localization data sets had four curation\nsteps including a manually built, comprehensive hierarchical structure of more\nthan 1600 subcellular localizations. ComPPI provides confidence scores for\nprotein subcellular localizations and protein-protein interactions. ComPPI has\nuser-friendly search options for individual proteins giving their subcellular\nlocalization, their interactions and the likelihood of their interactions\nconsidering the subcellular localization of their interacting partners.\nDownload options of search results, whole proteomes, organelle-specific\ninteractomes and subcellular localization data are available on its website.\nDue to its novel features, ComPPI is useful for the analysis of experimental\nresults in biochemistry and molecular biology, as well as for proteome-wide\nstudies in bioinformatics and network science helping cellular biology,\nmedicine and drug design.\n", " Making use of a unified approach to certain classes of induced\nrepresentations, we establish here a number of detailed spectral theoretic\ndecomposition results. They apply to specific problems from non-commutative\nharmonic analysis, ergodic theory, and dynamical systems. Our analysis is in\nthe setting of semidirect products, discrete subgroups, and solenoids. Our\napplications include analysis and ergodic theory of Bratteli diagrams and their\ncompact duals; of wavelet sets, and wavelet representations.\n", " The strong-field laser physics enterprise is investing important resources in\nthe study of the effects of oscillatory electric fields on matter using the\ntunneling concept, whereas laser fields are vector fields that do not support\nthe tunneling model. Oscillatory electric fields and propagating plane-wave\nlaser fields are different electrodynamic phenomena, and similarities in their\neffects diminish as field intensity increases. Major differences are known in\nthe case of very low frequencies where oscillatory electric fields act\nadiabatically as the frequency declines, in contrast to extreme relativistic\neffects of strong laser fields. Many supposed new phenomena, such as ATI\n(Above-Threshold Ionization), channel closing, and stabilization were studied\nin terms of propagating fields before they came to the attention of the atomic\nphysics community. This illustrates the efficiency of using proper\nelectrodynamic methods for the treatment of laser effects. Problems arising\nfrom the conflation of the effects of oscillatory electric fields and\npropagating fields are exacerbated by using ill-defined nomenclature, such as\nKFR and SFA, that has been used indiscriminately to apply both to oscillatory\nelectric fields and to propagating fields. Research resources can be applied\nwith much-improved efficiency if proper electrodynamic treatment of laser\nfields is employed.\n", " The advent of back-scattered laser beams for electron-positron colliders will\nallow detailed studies of a large array of high energy photon-photon and\nphoton-electron collision processes with polarized beams. These include tests\nof electroweak theory in photon-photon annihilation such as $\\gamma \\gamma \\to\nW^+ W^-$, $\\gamma \\gamma \\to $ Higgs bosons, and higher-order loop processes,\nsuch as $\\gamma \\gamma \\to \\gamma \\gamma, Z \\gamma, H^0 Z^0$ and $Z Z.$ Methods\nfor measuring the anomalous magnetic and quadrupole moments of the $W$ and $Z$\ngauge bosons to high precision in polarized electron-photon and photon-photon\ncollisions are discussed. Since each photon can be resolved into a $W^+ W^-$\npair, high energy photon-photon collisions can also provide a remarkably\nbackground-free laboratory for studying $W W$ collisions and annihilation. I\nalso review high energy $\\gamma \\gamma$ and $e \\gamma$ tests of quantum\nchromodynamics, including the production of two gluon jets in photon-photon\ncollisions, deeply virtual Compton scattering on a photon target, and\nleading-twist single-spin asymmetries for a photon polarized normal to a\nproduction plane. Exclusive hadron production processes in photon-photon\ncollisions provide important tests of QCD at the amplitude level, particularly\nas measures of hadron distribution amplitudes which are also important for the\nanalysis of exclusive semi-leptonic and two-body hadronic B-decays.\n", " The physical nature of the fluorescence depletion (FD) effect and the\npossible way of its experimental study are discussed.\n", " Stochastic kinetic models of genetic expression are able to describe protein\nfluctuations. A comparative study of the canonical and a feedback model is\ngiven here by using stochastic simulation methods. The feedback model is\nskeleton model implementation of the circular gene hypothesis, which suggests\nthe interaction between the synthesis and degradation of mRNA. Qualitative and\nquantitative changes in the shape and in the numerical characteristics of the\nstationary distributions suggest that more combined experimental and\ntheoretical studies should be done to uncover the details of the kinetic\nmechanism of gene expression.\n", " Folding of proteins into their correct native structure is key to their\nfunction. Simultaneously, the intricate interplay between cell movement and\nprotein conformation highlights the complex nature of cellular processes. In\nthis work, we demonstrate the impact of Terahertz (THz) signaling on\ncontrolling protein conformational changes in a random medium. Our system of\ninterest consists of a communication link that involves a nanoantenna\ntransmitter, a protein receiver, and a channel composed of moving red blood\ncells. Due to the system dynamics, we investigate the influence of both the\nfast and slow channel variations on protein folding. Specifically, we analyze\nthe system's selectivity to asses the effectiveness of the induced THz\ninteraction in targeting a specific group of proteins under fading conditions.\nBy optimizing the selectivity metric with respect to the nanoantenna power and\nfrequency, it is possible to enhance the controllability of protein\ninteractions. Our probabilistic analysis provides a new perspective regarding\nelectromagnetically triggered protein molecules, their microenvironment and\ntheir interaction with surrounding particles. It helps elucidate how external\nconditions impact the protein folding kinetics and pathways. This results in\nnot only understanding the mechanisms underlying THz-induced protein\ninteractions but also engineering these still-emerging tools.\n", " We used fluorescence spectroscopic measurements as a probe to study the\nself-healing properties of anthraquinone derivative molecules doped in\npoly(methyl methacrylate) (PMMA). 2,4-dibromo anthraquinone dye doped in PMMA\nrecovers after photodegradation. Its dynamics differs from other anthraquinone\nderivative molecules. This could be due to the relatively heavier bromine atom\nattached to one of the carbon atoms of the benzene ring. In this paper, we will\ndiscuss the self-healing properties of 2,4-dibromo anthraquinone doped in PMMA\nmatrix. We also tested the correlated chromophore domain model (CCrDM) and have\ncharacterized the self-healing properties by determining the CCrDM parameters.\nWe also estimated the self-absorption of fluorescence signal by the dye\nmolecule without which the analysis of the self-recovery of the molecule would\nbe incomplete.\n", " We consider the problem of inferring the unknown parameters of a stochastic\nbiochemical network model from a single measured time-course of the\nconcentration of some of the involved species. Such measurements are available,\ne.g., from live-cell fluorescence microscopy in image-based systems biology. In\naddition, fluctuation time-courses from, e.g., fluorescence correlation\nspectroscopy provide additional information about the system dynamics that can\nbe used to more robustly infer parameters than when considering only mean\nconcentrations. Estimating model parameters from a single experimental\ntrajectory enables single-cell measurements and quantification of cell--cell\nvariability. We propose a novel combination of an adaptive Monte Carlo sampler,\ncalled Gaussian Adaptation, and efficient exact stochastic simulation\nalgorithms that allows parameter identification from single stochastic\ntrajectories. We benchmark the proposed method on a linear and a non-linear\nreaction network at steady state and during transient phases. In addition, we\ndemonstrate that the present method also provides an ellipsoidal volume\nestimate of the viable part of parameter space and is able to estimate the\nphysical volume of the compartment in which the observed reactions take place.\n", " We investigate the phenomenological viability of real spin half, zero and one\ndark matter candidates, which interact predominantly with third generation\nheavy quarks and gluons via the twenty-eight gauge invariant higher dimensional\neffective operators. The corresponding Wilson coefficients are constrained one\nat a time from the relic density $\\Omega^{\\rm DM} h^2$ $\\approx$ 0.1198. Their\ncontributions to the thermal averaged annihilation cross-sections are shown to\nbe consistent with the FermiLAT and H.E.S.S. experiments' projected upper bound\non the annihilation cross-section in the $b\\,\\bar b$ mode.\n The tree-level gluon-philic and one-loop induced heavy-quark-philic\nDM-nucleon direct-detection cross-sections are analysed. The non-observation of\nany excess over expected background in the case of recoiled Xe-nucleus events\nfor spin-independent DM-nucleus scattering in XENON-1T sets the upper limits on\nthe eighteen Wilson coefficients. Our analysis validates the real DM candidates\nfor the large range of accessible mass spectrum below 2 TeV for all but one\ninteraction induced by the said operators.\n", " We describe a modification of the TAP method for purification and analysis of\nmultiprotein complexes, termed here DEF-TAP (for Differential Elution\nFractionation after Tandem Affinity Purification). Its essential new feature is\nthe use for last purification step of 6XHis-Ni++ interaction, which is\nresistant to a variety of harsh washing conditions, including high ionic\nstrength and presence of organic solvents. This allows us to use various\nfractionation schemes before the protease digestion, which is expected to\nimprove the coverage of the analysed protein mixture and also to provide an\nadditional insight into the structure of the purified macromolecular complex\nand the nature of protein-protein interactions involved. We illustrate our new\napproach by analysis of soluble nuclear complexes containing histone H4\npurified from HeLa cells. In particular, we observed different fractionation\npatterns of HAT1 and RbAp46 proteins as compared to RbAp48 protein, all\nidentified as interaction partners of H4 histone. In addition, we report all\ncomponents of the licensing MCM2-7 complex and the apoptosis-related DAXX\nprotein among the interaction partners of the soluble H4. Finally, we show that\nHAT1 requires N-terminal tail of H4 for its stable association with this\nhistone.\n", " The technique of laser ablation in liquids is applied to produce of\nBoron-containing nanoparticles from ablation of a Fe$_2$B bulk target enriched\nin 10B isotope. Laser ablation of the target in liquid isopropanol results in\npartial disproportionation to free Fe and Boron while nanoparticles of Fe$_2$B\nare also presented. Nanoparticles are magnetic and can be collected using a\npermanent magnet. Average size of nanoparticles is of 15 nm. The content of 10B\nin generated nanoparticles amounts to 76,9 %. Nanoparticles are biocompatible\nand can be used in Boron Neutron Capture Therapy.\n", " Torsion propagation and torsion-spin coupling are studied in the perspective\nof the Velo-Zwanziger method of analysis; specifically, we write the most\nextensive dynamics of the torsion tensor and the most exhaustive coupling that\nis permitted between torsion and spinors, and check the compatibility with\nconstraints and hyperbolicity and causality of field equations: we find that\nsome components of torsion and many terms of the torsion-spin interaction will\nbe restricted away and as a consequence we will present the most general theory\nthat is compatible with all restrictions.\n", " Biological networks provide insight into the complex organization of\nbiological processes in a cell at the system level. They are an effective tool\nfor understanding the comprehensive map of functional interactions, finding the\nfunctional modules and pathways. Reconstruction and comparative analysis of\nthese networks provide useful information to identify functional modules,\nprioritization of disease causing genes and also identification of drug\ntargets. The talk will consist of two parts. I will discuss several methods for\nprotein-protein interaction network alignment and investigate their preferences\nto other existing methods. Further, I briefly talk about reconstruction of\nprotein-protein interaction networks by using deep learning.\n", " We propose the mesoscopic device based on the Rashba spin orbit interaction\n(SOI) that contains a gated ballistic Aharonov-Bohm (AB) ring with incoming\nlead and two asymmetrically situated outgoing leads. The variations of the\nRashba coupling parameter induced by the gate voltage applied to the AB ring is\nshown to cause the redistribution of the carrier flux between the outgoing\nleads and spin polarization of the outgoing currents, thus allowing the system\nto manifest the properties of the quantum splitter and spin filter.\n", " Junctionless oxide-based neuron thin-film transistors with in-plane-gate\nstructure are fabricated at room temperature with a laser scribing process. The\nneuron transistors are composed of a bottom ITO floating gate and multiples of\ntwo in-plane control gates. The control gates, coupling with the floating gate,\ncontrol the \"on\" and \"off\" of the transistor. Effective field-effect modulation\nof the drain current has been realized. AND logic is demonstrated on a dual\nin-plane gate neuron transistor. The developed laser scribing technology is\nhighly desirable in terms of the fabrication of high performance neuron\ntransistors with low-cost.\n", " We consider emission of a photon by an electron in the field of a strong\nlaser wave. A probability of this process for circularly or linearly polarized\nlaser photons and for arbitrary polarization of all other particles is\ncalculated. We obtain the complete set of functions which describe such a\nprobability in a compact invariant form. Besides, we discuss in some detail the\npolarization effects in the kinematics relevant to the problem of electron to\nphoton conversion at photon-photon and electron-photon colliders.\n", " This report provides a brief review of recently developed extended framework\nfor fundamental physics, designated as Quantum Field Mechanics and including\ncausally complete and intrinsically unified theory of explicitly emerging\nelementary particles, their inherent properties, quantum and relativistic\nbehaviour, interactions and their results. Essential progress with respect to\nusual theory is attained due to the unreduced, nonperturbative analysis of\narbitrary interaction process revealing a qualitatively new phenomenon of\ndynamic multivaluedness of interaction results and their dynamically entangled\nstructure giving rise to universally defined dynamic complexity\n(physics/9806002) and absent in usual, always perturbative and dynamically\nsingle-valued models with the zero value of unreduced complexity (including\nsubstitutes used for \"complexity\", \"chaoticity\", \"self-organisation\", etc.). It\nis shown how the observed world structure, starting from elementary particles\nand their interactions, dynamically emerges from the unreduced interaction\nbetween two initially homogeneous, physically real protofields\n(quant-ph/9902015, quant-ph/9902016). In that way one avoids arbitrary\nimposition of abstract entities and unprovable postulates or \"principles\", let\nalone \"mysteries\", of conventional theory, which now obtain unified, consistent\nand realistic explanation in terms of unreduced dynamic complexity. We complete\nthe description of fundamental world structure emergence and properties by an\noutline of experimental implications and resulting causally substantiated\nchange of research strategy.\n", " Van der Waals forces as interactions between neutral and polarisable\nparticles act at small distances between two objects. Their theoretical origin\nlies in the electromagnetic interaction between induced dipole moments caused\nby the vacuum fluctuations of the ground-state electromagnetic field. The\nresulting theory well describes the experimental situation in the limit of the\npoint dipole assumption. At smaller distances, where the finite size of the\nparticles has to be taken into account, this description fails and has to be\ncorrected by higher orders of the multipole expansion, such as quadrupole\nmoments and so on. With respect to the complexity of the spatial properties of\nthe particles this task requires a considerable effort. In order to describe\nthe van der Waals interaction between such particles, we apply the established\nmethod of a spatially spread out polarisability distribution to approximate the\nhigher orders of the multipole expansion. We hence construct an effective\ntheory for effects from anisotropy and finite size on the van der Waals\npotential.\n", " The capability of fewest-switches surface hopping (FSSH) to describe\nnon-adiabatic dynamics of small and medium sized molecules under explicit\nexcitation with external fields is evaluated. Different parameters in FSSH and\ncombinations thereof are benchmarked against multi-configurational time\ndependent Hartree (MCTDH) reference calculations using SO$_2$ and\n2-thiocytosine as model, yet realistic, molecular systems. Qualitatively, we\nfind that FSSH is able to reproduce the trends in the MCTDH dynamics with (and\nwithout) an explicit external field; however, no set of FSSH parameters is\nideal. An adequate treatment of the overcoherence in FSSH is identified as the\ndriving factor to improve the description of the excitation process with\nrespect to the MCTDH reference. Here two corrections were tested, the\naugmented-FSSH (AFSSH) and the energy-based decoherence correction. A\ndependence on the employed basis is detected for the AFSSH algorithm,\nperforming better when spin-orbit and external laser field couplings are\ntreated as off-diagonal elements instead of projecting them onto the diagonal\nof the Hamilton operator. In the presence of an electric field, the excited\nstate dynamics was found to depend strongly on the vector used to rescale the\nkinetic energy along after a transition between surfaces. For SO$_2$,\nrecurrence of the excited wave packet throughout the duration of the applied\nlaser pulse is observed for long laser pulses (>100~fs), resulting in\nadditional interferences not captured by FSSH and only visible in variational\nmulti-configurational Gaussian when utilizing a large amount of gaussian basis\nfunctions. This feature essentially vanishes when going towards larger\nmolecules, such as 2-thiocytosine, where this effect is barely visible in a\nlaser pulse with a full width at half maximum of 200~fs.\n", " The LiFr diatomic represents a promising candidate for indirect laser cooling\nthat has not yet been investigated not theoretically or experimentally. The\npotential energy curves of the ground and low_lying excited states of the LiFr\nheteronuclear alkali metal dimer are calculated using the Fock_space\nrelativistic coupled cluster theory for the first time. A number of properties\nsuch as the electronic term energies, equilibrium internuclear distances,\ntransition and permanent dipole moments, sequences of vibrational energies,\nharmonic vibrational frequencies, Franck_Condon factors, and radiative\nlifetimes (including bound and free transitions) are predicted. The\nprobabilities of the two_step schemes (optical cycles) for the transfer process\nof the LiFr molecules from high excited vibrational states to the ground\nvibronic state are also predicted. The data obtained would be useful for laser\ncooling and spectral experiments with LiFr molecules.\n", " An analysis of known experiments on direct and indirect detection of the\nelectron-type neutrino and antineutrino has been made. The analysis based on a\nnew hypothesis that the observed physical space is formed from a finite set of\nbyuons, \"one-dimensional vectorial objects\". It is shown in the article that\nthe hypothesis for existence of neutrinos advanced by Pauli on the basis of an\nanalysis of the conservation laws, is not unquestionable since the fulfillment\nof these laws may be secured by the physical space itself (physical vacuum)\nbeing the lowest energy state of a discrete oscillating system originating in\nthe course of byuon interaction. This effect is analogous to that of\nM\\\"ossbauer. The direct experiments on detecting neutrinos are explained from\nthe existence of a new information channel due to the uncertainty interval for\ncoordinate of the four-contact byuon interaction forming the interior geometry\nof elementary particles and their properties. Given are also the results of an\nexperiment on observation of cyclic variations of the $\\beta$-decay rate, which\nconfirm the existence of said new information channel.\n", " We intercalate a van der Waals heterostructure of graphene and hexagonal\nBoron Nitride with Au, by encapsulation, and show that Au at the interface is\ntwo dimensional. A charge transfer upon current annealing indicates\nredistribution of Au and induces splitting of the graphene bandstructure. The\neffect of in plane magnetic field confirms that splitting is due to\nspin-splitting and that spin polarization is in the plane, characteristic of a\nRashba interaction with magnitude approximately 25 meV. Consistent with the\npresence of intrinsic interfacial electric field we show that the splitting can\nbe enhanced by an applied displacement field in dual gated samples. Giant\nnegative magnetoresistance, up to 75%, and a field induced anomalous Hall\neffect at magnetic fields < 1 T are observed. These demonstrate that hybridized\nAu has a magnetic moment and suggests the proximity to formation of a\ncollective magnetic phase. These effects persist close to room temperature.\n", " Universe structure emerges in the unreduced, complex-dynamic interaction\nprocess with the simplest initial configuration (two attracting homogeneous\nfields, quant-ph/9902015). The unreduced interaction analysis gives\nintrinsically creative cosmology, describing the real, explicitly emerging\nworld structure with dynamic randomness on each scale. Without imposing any\npostulates or entities, we obtain physically real space, time, elementary\nparticles with their detailed structure and intrinsic properties, causally\ncomplete and unified version of quantum and relativistic behaviour, the origin\nand number of naturally unified fundamental forces, and classical behaviour\nemergence in closed systems (gr-qc/9906077). Main problems of standard\ncosmology and astrophysics are consistently solved in this extended picture\n(without introduction of any additional entities), including those of quantum\ncosmology and gravity, entropy growth and time, \"hierarchy\" of elementary\nparticles, \"anthropic\" difficulties, space-time flatness, and \"missing\"\n(\"dark\") mass and energy. The observed universe structure and laws can be\npresented as manifestations of the universal symmetry (conservation) of\ncomplexity providing the unified, irregular, but exact (never \"broken\") Order\nof the World (physics/0404006).\n", " In recent years, the question of the reliability of Machine Learning (ML)\nmethods has acquired significant importance, and the analysis of the associated\nuncertainties has motivated a growing amount of research. However, most of\nthese studies have applied standard error analysis to ML models, and in\nparticular Deep Neural Network (DNN) models, which represent a rather\nsignificant departure from standard scientific modelling. It is therefore\nnecessary to integrate the standard error analysis with a deeper\nepistemological analysis of the possible differences between DNN models and\nstandard scientific modelling and the possible implications of these\ndifferences in the assessment of reliability. This article offers several\ncontributions. First, it emphasises the ubiquitous role of model assumptions\n(both in ML and traditional Science) against the illusion of theory-free\nscience. Secondly, model assumptions are analysed from the point of view of\ntheir (epistemic) complexity, which is shown to be language-independent. It is\nargued that the high epistemic complexity of DNN models hinders the estimate of\ntheir reliability and also their prospect of long-term progress. Some potential\nways forward are suggested. Thirdly, this article identifies the close relation\nbetween a model's epistemic complexity and its interpretability, as introduced\nin the context of responsible AI. This clarifies in which sense, and to what\nextent, the lack of understanding of a model (black-box problem) impacts its\ninterpretability in a way that is independent of individual skills. It also\nclarifies how interpretability is a precondition for assessing the reliability\nof any model, which cannot be based on statistical analysis alone. This article\nfocuses on the comparison between traditional scientific models and DNN models.\nBut, Random Forest and Logistic Regression models are also briefly considered.\n", " This thesis consists of two topics related to computational geometry and one\ntopic related to topological data analysis (TDA), which combines fields of\ncomputational geometry and algebraic topology for analyzing data. The first\npart studies the classical problem of finding k nearest neighbors to m query\npoints in a larger set of n reference points in any metric space. The second\npart is about the construction of a Minimum Spanning Tree (MST) on any finite\nmetric space. The third part extends the key concept of persistence within\nTopological Data Analysis in a new direction.\n", " The 0-instanton solution of Painlev\\'e I is a sequence $(u_{n,0})$ of complex\nnumbers which appears universally in many enumerative problems in algebraic\ngeometry, graph theory, matrix models and 2-dimensional quantum gravity. The\nasymptotics of the 0-instanton $(u_{n,0})$ for large $n$ were obtained by the\nthird author using the Riemann-Hilbert approach. For $k=0,1,2,...$, the\n$k$-instanton solution of Painlev\\'e I is a doubly-indexed sequence $(u_{n,k})$\nof complex numbers that satisfies an explicit quadratic non-linear recursion\nrelation. The goal of the paper is three-fold: (a) to compute the asymptotics\nof the 1-instanton sequence $(u_{n,1})$ to all orders in $1/n$ by using the\nRiemann-Hilbert method, (b) to present formulas for the asymptotics of\n$(u_{n,k})$ for fixed $k$ and to all orders in $1/n$ using resurgent analysis,\nand (c) to confirm numerically the predictions of resurgent analysis. We point\nout that the instanton solutions display a new type of Stokes behavior, induced\nfrom the tritronqu\\'ee Painlev\\'e transcendents, and which we call the induced\nStokes phenomenon. The asymptotics of the 2-instanton and beyond exhibits new\nphenomena not seen in 0 and 1-instantons, and their enumerative context is at\npresent unknown.\n", " In this paper, we study merge trees induced by a discrete Morse function on a\ntree. Given a discrete Morse function, we provide a method to constructing an\ninduced merge tree and define a new notion of equivalence of discrete Morse\nfunctions based on the induced merge tree. We then relate the matching number\nof a tree to a certain invariant of the induced merge tree. Finally, we count\nthe number of merge trees that can be induced on a star graph and characterize\nthe induced merge tree.\n", " The sensitivity of the displacement transducer pumped with a train of\nhigh-intensity laser pulses is estimated. Due to the multicomponent character\nof the pump a consideration of transformations of the signal and the noises\nbetween optical modes plays an important role in estimation of the potential\nsensitivity. An expression for the minimal detectable external classical force\nresembles those for the continuous wave pumping with substitution of the laser\npower by a time averaged power of pulsed laser. Possible scheme for back action\nnoise compensation for such transducers is considered. For full suppression of\nback action noise the field of local oscillator has to be pulsed with the same\ntime dependence as the pump field.\n", " Local perturbations of individual metabolic reactions may result in different\nlevels of lethality, depending on their roles in metabolism and the size of\nsubsequent cascades induced by their failure. Moreover, essentiality of\nindividual metabolic reactions may show large variations within and across\nspecies. Here we quantify their essentialities in hundreds of species by\ncomputing the growth rate after removal of individual and pairs of reactions by\nflux balance analysis. We find that about 10% of reactions are essential, i.e.,\ngrowth stops without them, and most of the remaining reactions are redundant in\nthe metabolic network of each species. This large-scale and cross-species study\nallows us to determine ad hoc ages of each reaction and species. We find that\nwhen a reaction is older and contained in younger species, the reaction is more\nlikely to be essential. Such correlations of essentiality with the ages of\nreactions and species may be attributable to the evolution of cellular\nmetabolism, in which alternative pathways are recruited to ensure the stability\nof important reactions to various degrees across species.\n", " Ultrafast non-thermal manipulation of magnetization by light relies on either\nindirect coupling of the electric field component of the light with spins via\nspin-orbit interaction or direct coupling between the magnetic field component\nand spins. Here we propose a novel scenario for coupling between the electric\nfield of light and spins via optical modification of the exchange interaction,\none of the strongest quantum effects, the strength of which can reach 1000\nTesla. We demonstrate that this isotropic opto-magnetic effect, which can be\ncalled the inverse magneto-refraction, is allowed in a material of any\nsymmetry. Its existence is corroborated by the experimental observation of THz\nemission by magnetic-dipole active spin resonances optically excited in a broad\nclass of iron oxides with a canted spin configuration. From its strength we\nestimate that a sub-picosecond laser pulse with a moderate fluence of ~ 1\nmJ/cm^2 acts as a pulsed effective magnetic field of 0.01 Tesla, arising from\nthe optically perturbed balance between the exchange parameters. Our findings\nare supported by a low-energy theory for the microscopic magnetic interactions\nbetween non-equilibrium electrons subjected to an optical field which suggests\na possibility to modify the exchange interactions by light over 1 %.\n", " The goal of the present article is to review the major developments that have\nled to the current understanding of molecule-field interactions and\nexperimental methods for manipulating molecules with electromagnetic fields.\nMolecule-field interactions are at the core of several, seemingly distinct,\nareas of molecular physics. This is reflected in the organization of this\narticle, which includes sections on Field control of molecular beams, External\nfield traps for cold molecules, Control of molecular orientation and molecular\nalignment, Manipulation of molecules by non-conservative forces, Ultracold\nmolecules and ultracold chemistry, Controlled many-body phenomena, Entanglement\nof molecules and dipole arrays, and Stability of molecular systems in\nhigh-frequency super-intense laser fields. The article contains 853 references.\n", " Graph convexity spaces have been studied in many contexts. In particular,\nsome studies are devoted to determine if a graph equipped with a convexity\nspace is a {\\em convex geometry}. It is well known that chordal and Ptolemaic\ngraphs can be characterized as convex geometries with respect to the geodesic\nand monophonic convexities, respectively. Weak polarizable graphs, interval\ngraphs, and proper interval graphs can also be characterized in this way. In\nthis paper we introduce the notion of {\\em $l^k$-convexity}, a natural\nrestriction of the monophonic convexity. Let $G$ be a graph and $k\\geq 2$ an\ninteger. A subset $S\\subseteq V(G)$ is \\textit{$l^k$-convex} if and only if for\nany pair of vertices $x,y$ of $S$, each induced path of length {\\em at most}\n$k$ connecting $x$ and $y$ is completely contained in the subgraph induced by\n$S$. The {\\em $l^k$-convexity} consists of all $l^k$-convex subsets of $G$. In\nthis work, we characterize {\\em $l^k$-convex geometries} (graphs that are\nconvex geometries with respect to the $l^k$-convexity) for $k\\in\\{2,3\\}$. We\nshow that a graph $G$ is an $l^2$-convex geometry if and only if $G$ is a\nchordal $P_4$-free graph, and an $l^3$-convex geometry if and only if $G$ is a\nchordal graph with diameter at most three such that its induced gems satisfy a\nspecial \"solving\" property. As far as the authors know, the class of\n$l^3$-convex geometries is the first example of a non-hereditary class of\nconvex geometries.\n", " It is obvious that both epigenetic and non-epigenetic actors contribute to\ntumorigenesis in chondrosarcomas and more generally in other cancers. Thus, the\nmain altered pathways in chondrosarcomas are now well established and include\nboth epigenetic and non-epigenetic pathways such as the PI3K-AKT signaling,\nEGFR overexpression, SPARC overexpression, c-myc overexpression, IHH/GLI1 axis,\nloss of Rb function, HIF1-alpha stabilization, IDH1 mutations, hypermethylation\nand SIRT1. This review aims to provide a detailed analysis of these pathways\nand highlights recurrent interactions between non-epigenetic and epigenetic\nactors in chondrosarcomas, raising the intriguing possibility of developing\ntherapeutics targeting both epigenetic and non-epigenetic actors and supporting\ndata from previous studies. Finally, we propose some strategies for targeting\nchondrosarcomas in vivo based on properties of this tumor.\n", " The walls of the box which contains matter represent a membrane that allows\nthe relativistic quantum vacuum to pass but not matter. That is why the\npressure of matter in the box may be considered as the analog of the osmotic\npressure. However, we demonstrate that the osmotic pressure of matter is\nmodified due to interaction of matter with vacuum. This interaction induces the\nnonzero negative vacuum pressure inside the box, as a result the measured\nosmotic pressure becomes smaller than the matter pressure. As distinct from the\nCasimir effect, this induced vacuum pressure is the bulk effect and does not\ndepend on the size of the box. This effect dominates in the thermodynamic limit\nof the infinite volume of the box. Analog of this effect has been observed in\nthe dilute solution of 3He in liquid 4He, where the superfluid 4He plays the\nrole of the non-relativistic quantum vacuum, and 3He atoms play the role of\nmatter.\n", " In this paper we propose network methodology to infer prognostic cancer\nbiomarkers based on the epigenetic pattern DNA methylation. Epigenetic\nprocesses such as DNA methylation reflect environmental risk factors, and are\nincreasingly recognised for their fundamental role in diseases such as cancer.\nDNA methylation is a gene-regulatory pattern, and hence provides a means by\nwhich to assess genomic regulatory interactions. Network models are a natural\nway to represent and analyse groups of such interactions. The utility of\nnetwork models also increases as the quantity of data and number of variables\nincrease, making them increasingly relevant to large-scale genomic studies. We\npropose methodology to infer prognostic genomic networks from a DNA\nmethylation-based measure of genomic interaction and association. We then show\nhow to identify prognostic biomarkers from such networks, which we term\n`network community oncomarkers'. We illustrate the power of our proposed\nmethodology in the context of a large publicly available breast cancer dataset.\n", " A dynamic logic method was developed to analyze molecular networks of cells\nby combining Kauffman and Thomas's logic operations with molecular interaction\nparameters. The logic operations characterize the discrete interactions between\nbiological components. The interaction parameters (e.g. response times)\ndescribe the quantitative kinetics. The combination of the two quantitatively\ncharacterizes the discrete biological interactions. A number of simple networks\nwere analyzed. The main results include: we proved the theorems to determine\nbistable states and oscillation behaviors of networks, we showed that time\ndelays are essential for oscillation structures, we proved that single variable\nnetworks do not have chaotic behaviors, and we explained why one signal can\nhave multiply responses. In addition, we applied the present method to the\nanalysis of the MAPK cascade, feed-forward loops, and mitosis cycle of budding\nyeast cells.\n", " Given the existing empirical information about the exotic Theta+ baryon, we\nanalyze possible properties of its SU(3)F-partners, paying special attention to\nthe nonstrange member of the antidecuplet N*. The modified piN partial-wave\nanalysis presents two candidate masses, 1680 MeV and 1730 MeV. In both cases,\nthe N* should be rather narrow and highly inelastic. Our results suggest\nseveral directions for experimental studies that may clarify properties of the\nantidecuplet baryons, and structure of their mixing with other baryons. Recent\nexperimental evidence from the GRAAL and STAR Collaborations could be\ninterpreted as observations of a candidate for the Theta+ nonstrange partner.\n" ]

[ " Given a set S of n \\geq d points in general position in R^d, a random\nhyperplane split is obtained by sampling d points uniformly at random without\nreplacement from S and splitting based on their affine hull. A random\nhyperplane search tree is a binary space partition tree obtained by recursive\napplication of random hyperplane splits. We investigate the structural\ndistributions of such random trees with a particular focus on the growth with\nd. A blessing of dimensionality arises--as d increases, random hyperplane\nsplits more closely resemble perfectly balanced splits; in turn, random\nhyperplane search trees more closely resemble perfectly balanced binary search\ntrees.\n We prove that, for any fixed dimension d, a random hyperplane search tree\nstoring n points has height at most (1 + O(1/sqrt(d))) log_2 n and average\nelement depth at most (1 + O(1/d)) log_2 n with high probability as n\n\\rightarrow \\infty. Further, we show that these bounds are asymptotically\noptimal with respect to d.\n", " In this paper we propose two different primal-dual splitting algorithms for\nsolving inclusions involving mixtures of composite and parallel-sum type\nmonotone operators which rely on an inexact Douglas-Rachford splitting method,\nhowever applied in different underlying Hilbert spaces. Most importantly, the\nalgorithms allow to process the bounded linear operators and the set-valued\noperators occurring in the formulation of the monotone inclusion problem\nseparately at each iteration, the latter being individually accessed via their\nresolvents. The performances of the primal-dual algorithms are emphasized via\nsome numerical experiments on location and image deblurring problems.\n", " We show that the set of binary words containing overlaps is not unambiguously\ncontext-free and that the set of ternary words containing overlaps is not\ncontext-free. We also show that the set of binary words that are not subwords\nof the Thue-Morse word is not unambiguously context-free.\n", " An alternative proof is given of the existence of greatest lower bounds in\nthe imbalance order of binary maximal instantaneous codes of a given size.\nThese codes are viewed as maximal antichains of a given size in the infinite\nbinary tree of 0-1 words. The proof proposed makes use of a single balancing\noperation instead of expansion and contraction as in the original proof of the\nexistence of glb.\n", " We consider \\textit{additive spaces}, consisting of two intervals of unit\nlength or two general probability measures on ${\\mathbb R}^1$, positioned on\nthe axes in ${\\mathbb R}^2$, with a natural additive measure $\\rho$. We study\nthe relationship between the exponential frames, Riesz bases, and orthonormal\nbases of $L^2(\\rho)$ and those of its component spaces. We find that the\nexistence of exponential bases depends strongly on how we position our measures\non ${\\mathbb R}^1$. We show that non-overlapping additive spaces possess Riesz\nbases, and we give a necessary condition for overlapping spaces. We also show\nthat some overlapping additive spaces of Lebesgue type have exponential\northonormal bases, while some do not. A particular example is the \"L\" shape at\nthe origin, which has a unique orthonormal basis up to translations of the form\n\\[ \\left\\{e^{2 \\pi i (\\lambda_1 x_1 + \\lambda_2 x_2)} : (\\lambda_1, \\lambda_2)\n\\in \\Lambda \\right\\}, \\] where \\[ \\Lambda = \\{ (n/2, -n/2) \\mid n \\in {\\mathbb\nZ} \\}. \\]\n", " We derive formulas for $(i)$ the number of toroidal $n\\times n$ binary\narrays, allowing rotation of rows and/or columns as well as matrix\ntransposition, and $(ii)$ the number of toroidal $n\\times n$ binary arrays,\nallowing rotation and/or reflection of rows and/or columns as well as matrix\ntransposition.\n", " We show that the number of prefix normal binary words of length $n$ is\n$2^{n-\\Theta((\\log n)^2)}$. We also show that the maximum number of binary\nwords of length $n$ with a given fixed prefix normal form is\n$2^{n-O(\\sqrt{n\\log n})}$.\n", " Some additive reverses of the continuous triangle inequality for Bochner\nintegral of vector-valued functions in Hilbert spaces are given. Applications\nfor complex-valued functions are provided as well.\n", " Dispersion, impedance matching and resolution characteristics of an isotropic\nthree-dimensional flat lens (\"superlens\") are studied. The lens is based on\ncubic meshes of interconnected transmission lines and bulk loads. We study a\npractical realization of the lens, based on the microstrip technology. The\ndispersion equations that have been previously derived, are verified with\nfull-wave simulations. The isotropy of the structure is verified with\nanalytical as well as simulation results. The resolution characteristics of a\npractically realizable, lossy lens are studied analytically.\n", " Physical phenomena in science and engineering are frequently modeled using\nscalar fields. In scalar field topology, graph-based topological descriptors\nsuch as merge trees, contour trees, and Reeb graphs are commonly used to\ncharacterize topological changes in the (sub)level sets of scalar fields. One\nof the biggest challenges and opportunities to advance topology-based\nvisualization is to understand and incorporate uncertainty into such\ntopological descriptors to effectively reason about their underlying data. In\nthis paper, we study a structural average of a set of labeled merge trees and\nuse it to encode uncertainty in data. Specifically, we compute a 1-center tree\nthat minimizes its maximum distance to any other tree in the set under a\nwell-defined metric called the interleaving distance. We provide heuristic\nstrategies that compute structural averages of merge trees whose labels do not\nfully agree. We further provide an interactive visualization system that\nresembles a numerical calculator that takes as input a set of merge trees and\noutputs a tree as their structural average. We also highlight structural\nsimilarities between the input and the average and incorporate uncertainty\ninformation for visual exploration. We develop a novel measure of uncertainty,\nreferred to as consistency, via a metric-space view of the input trees.\nFinally, we demonstrate an application of our framework through merge trees\nthat arise from ensembles of scalar fields. Our work is the first to employ\ninterleaving distances and consistency to study a global, mathematically\nrigorous, structural average of merge trees in the context of uncertainty\nvisualization.\n", " We model biochemical signal transduction, based on a ligand-receptor binding\nmechanism, as a discrete-time finite-state Markov channel, which we call the\nBIND channel. We show how to obtain the capacity of this channel, for the case\nof binary output, binary channel state, and arbitrary finite input alphabets.\nWe show that the capacity-achieving input distribution is IID. Further, we show\nthat feedback does not increase the capacity of this channel. We show how the\ncapacity of the discrete-time channel approaches the capacity of Kabanov's\nPoisson channel, in the limit of short time steps and rapid ligand release.\n", " Discrete geometric estimators aim at estimating geometric characteristics of\na shape with only its digitization as input data. Such an estimator is\nmultigrid convergent when its estimates tend toward the geometric\ncharacteristics of the shape as the digitization step h tends toward 0. This\npaper studies the multigrid convergence of tangent estimators based on maximal\ndigital straight segment recognition. We show that such estimators are\nmultigrid convergent for some family of convex shapes and that their speed of\nconvergence is on average O(h^(2/3)). Experiments confirm this result and\nsuggest that the bound is tight.\n", " The kinematics laws of electrons motion in the volume of a planar vacuum\ndiode running in regime of current saturation are used. The characteristics of\ndiode hysteresis in the conditions of S-figurative instability are got and\nanalyzed.\n", " Pippenger ([Pippenger, 1977]) showed the existence of\n$(6m,4m,3m,6)$-concentrator for each positive integer $m$ using a probabilistic\nmethod. We generalize his approach and prove existence of\n$(6m,4m,3m,5.05)$-concentrator (which is no longer regular, but has fewer\nedges). We apply this result to improve the constant of approximation of almost\nadditive set functions by additive set functions from $44.5$ (established in\n[Kalton, Roberts, 1983]) to $39$. We show a more direct connection of the\nlatter problem to the Whitney type estimate for approximation of continuous\nfunctions on a cube in $\\mathbb{R}^d$ by linear functions, and improve the\nestimate of this Whitney constant from $802$ (proved in [Brudnyi, Kalton,\n2000]) to $73$.\n", " In this paper we study the main characteristics of some evaluation codes\nparameterized by the edges of a bipartite graph with a perfect matching.\n", " In many complex molecular systems, the macroscopic ensemble's properties are\ncontrolled by microscopic dynamic events (or fluctuations) that are often\ndifficult to detect via pattern-recognition approaches. Discovering the\nrelationships between local structural environments and the dynamical events\noriginating from them would allow unveiling microscopic level\nstructure-dynamics relationships fundamental to understand the macroscopic\nbehavior of complex systems. Here we show that, by coupling advanced structural\n(e.g., Smooth Overlap of Atomic Positions, SOAP) with local dynamical\ndescriptors (e.g., Local Environment and Neighbor Shuffling, LENS) in a unique\ndataset, it is possible to improve both individual SOAP- and LENS-based\nanalyses, obtaining a more complete characterization of the system under study.\nAs representative examples, we use various molecular systems with diverse\ninternal structural dynamics. On the one hand, we demonstrate how the\ncombination of structural and dynamical descriptors facilitates decoupling\nrelevant dynamical fluctuations from noise, overcoming the intrinsic limits of\nthe individual analyses. Furthermore, machine learning approaches also allow\nextracting from such combined structural/dynamical dataset useful\nmicroscopic-level relationships, relating key local dynamical events (e.g.,\nLENS fluctuations) occurring in the systems to the local structural (SOAP)\nenvironments they originate from. Given its abstract nature, we believe that\nsuch an approach will be useful in revealing hidden microscopic\nstructure-dynamics relationships fundamental to rationalize the behavior of a\nvariety of complex systems, not necessarily limited to the atomistic and\nmolecular scales.\n", " This research article compares two characteristics of solutions of linear\ndifferential equations of the third order with variable coefficients. It\nappears that there is a lower estimate for wandering rate of solution to a\nlinear equation in terms of its frequency.\n", " The Chou-Yang model based on multiple diffraction theory has already been\nused by us to predict the characteristics of antiproton-proton elastic\nscattering at very high energies. But the recent precise data on proton elastic\nform factor obtained at JLab essentially deviates from the dipole\napproximation. We have parametrized the data for the form factor. The\ncomputations have been made by using the new expression for the form factor.\nThe implications of the new form factor on the characteristics of\nantiproton-proton elastic scattering at very high energy have been studied.\n", " We investigated Capacitance-Voltage (C-V) characteristics of the Depletion\nMode Buried Channel InGaAs/InAs Quantum Well FET by using Self-Consistent\nmethod incorporating Quantum Mechanical (QM) effects. Though the experimental\nresults of C-V for enhancement type device is available in recent literature, a\ncomplete characterization of electrostatic property of depletion type Buried\nChannel Quantum Well FET (QWFET) structure is yet to be done. C-V\ncharacteristics of the device is studied with the variation of three important\nprocess parameters: Indium (In) composition, gate dielectric and oxide\nthickness. We observed that inversion capacitance and ballistic current tend to\nincrease with the increase in Indium (In) content in InGaAs barrier layer.\n", " CG:SHOP is an annual geometric optimization challenge and the 2022 edition\nproposed the problem of coloring a certain geometric graph defined by line\nsegments. Surprisingly, the top three teams used the same technique, called\nconflict optimization. This technique has been introduced in the 2021 edition\nof the challenge, to solve a coordinated motion planning problem. In this\npaper, we present the technique in the more general framework of binary\nconstraint satisfaction problems (binary CSP). Then, the top three teams\ndescribe their different implementations of the same underlying strategy. We\nevaluate the performance of those implementations to vertex color not only\ngeometric graphs, but also other types of graphs.\n", " We start by considering binary words containing the minimum possible numbers\nof squares and antisquares (where an antisquare is a word of the form $x\n\\overline{x}$), and we completely classify which possibilities can occur. We\nconsider avoiding $x p(x)$, where $p$ is any permutation of the underlying\nalphabet, and $x t(x)$, where $t$ is any transformation of the underlying\nalphabet. Finally, we prove the existence of an infinite binary word\nsimultaneously avoiding all occurrences of $x h(x)$ for every nonerasing\nmorphism $h$ and all sufficiently large words $x$.\n", " Ivan Dobrakov has initiated a theory of non-additive set functions defined on\na ring of sets intended to be a non-additive generalization of the theory of\nfinite non-negative countably additive measures. These set functions are now\nknown as the Dobrakov submeasures. In this paper we extend Dobrakov's\nconsiderations to vector-valued submeasures defined on a ring of sets. The\nextension of such submeasures in the sense of Drewnowski is also given.\n", " Evidence is summarized attesting that the standard exchange field theory of\nferromagnetism by Heisenberg has not been successful. It is replaced by the\ncrystal field and a simple assumption that spin orientation is inexorably\nassociated with the orientation of its carrier. It follows at once that both\nferromagnetic phase transitions and magnetization must involve a structural\nrearrangement. The mechanism of structural rearrangements in solids is\nnucleation and interface propagation. The new approach accounts coherently for\nferromagnetic state and its manifestations.\n", " We investigate the collision cascade that is generated by a single moving\nincident particle on a static hard-sphere gas. We argue that the number of\nmoving particles at time t grows as t^{xi} and the number collisions up to time\nt grows as t^{eta}, with xi=2d/(d+2) and eta=2(d+1)/(d+2) and d the spatial\ndimension. These growth laws are the same as those from a hydrodynamic theory\nfor the shock wave emanating from an explosion. Our predictions are verified by\nmolecular dynamics simulations in d=1 and 2. For a particle incident on a\nstatic gas in a half-space, the resulting backsplatter ultimately contains\nalmost all the initial energy.\n", " We investigate the possible sequels, for the hybrid compact stars consisting\nof nucleons, hyperons and three flavour color-flavor-locked quarks phase under\nglobal neutrality and chemical equilibrium conditions. The hadron equations of\nstate are computed within the framework of energy density functionals based on\nthe relativistic mean field theory by employing two different model\nparameterizations. The quarks matter phase of equation of state is computed by\nusing Quarks Quasiparticle model derived from a non-relativistic energy\ndensity-functional approach. A plausible set of hybrid equations of state for\nsuperdense hadron-quarks matter is obtained and, which satisfies the\nconstraints provided by bulk nuclear matter, the observational data of\nfrequencies and maximum Gravitational mass 2$M_\\odot$ with their extracted\nradii, Keplerian limits, secular axisymmetric instability and limits of tidal\ndeformabilities from GW170817 binary neutron stars (BNS) merger. We obtained\nstructural properties of nonrotating and rotating compact stars, the\nevolutionary sequences with constant baryonic mass star spinning down by\nelectromagnetic and gravitational radiations and, discuss the issues relating\nto phase transition from hadrons to quarks matter phase. The internal structure\nof rotating star with observed spin down frequencies, exhibiting shrinkage of\nsoft quarks core of compact stars are discussed for constant baryonic mass. We\npresent the theoretically computed limits of radii for the spin down\nconfigurations of hybrid stars corresponding to the recently observed\nmillisecond pulsars. We also present and discuss the calculated results for\ntidal deformability, $\\Lambda$ and tidal parameters $\\Lambda_1$, $\\Lambda_2$ of\nnon spining hybrid star which are consistent with the waveform model analysis\nof GW170817 BNS.\n", " Graph Neural Networks (GNNs) have emerged as a powerful and flexible\nframework for representation learning on irregular data. As they generalize the\noperations of classical CNNs on grids to arbitrary topologies, GNNs also bring\nmuch of the implementation challenges of their Euclidean counterparts. Model\nsize, memory footprint, and energy consumption are common concerns for many\nreal-world applications. Network binarization allocates a single bit to\nparameters and activations, thus dramatically reducing the memory requirements\n(up to 32x compared to single-precision floating-point numbers) and maximizing\nthe benefits of fast SIMD instructions on modern hardware for measurable\nspeedups. However, in spite of the large body of work on binarization for\nclassical CNNs, this area remains largely unexplored in geometric deep\nlearning. In this paper, we present and evaluate different strategies for the\nbinarization of graph neural networks. We show that through careful design of\nthe models, and control of the training process, binary graph neural networks\ncan be trained at only a moderate cost in accuracy on challenging benchmarks.\nIn particular, we present the first dynamic graph neural network in Hamming\nspace, able to leverage efficient k-NN search on binary vectors to speed-up the\nconstruction of the dynamic graph. We further verify that the binary models\noffer significant savings on embedded devices. Our code is publicly available\non Github.\n", " Protein representation learning methods have shown great potential to yield\nuseful representation for many downstream tasks, especially on protein\nclassification. Moreover, a few recent studies have shown great promise in\naddressing insufficient labels of proteins with self-supervised learning\nmethods. However, existing protein language models are usually pretrained on\nprotein sequences without considering the important protein structural\ninformation. To this end, we propose a novel structure-aware protein\nself-supervised learning method to effectively capture structural information\nof proteins. In particular, a well-designed graph neural network (GNN) model is\npretrained to preserve the protein structural information with self-supervised\ntasks from a pairwise residue distance perspective and a dihedral angle\nperspective, respectively. Furthermore, we propose to leverage the available\nprotein language model pretrained on protein sequences to enhance the\nself-supervised learning. Specifically, we identify the relation between the\nsequential information in the protein language model and the structural\ninformation in the specially designed GNN model via a novel pseudo bi-level\noptimization scheme. Experiments on several supervised downstream tasks verify\nthe effectiveness of our proposed method.The code of the proposed method is\navailable in \\url{https://github.com/GGchen1997/STEPS_Bioinformatics}.\n", " In this paper we propose an algorithm to generate binary words with no more\n0's than 1's having a fixed number of 1's and avoiding the pattern $(10)^j1$\nfor any fixed $j \\geq 1$. We will prove that this generation is exhaustive,\nthat is, all such binary words are generated.\n", " Drug development is a lengthy process with a high failure rate. Increasingly,\nmachine learning is utilized to facilitate the drug development processes.\nThese models aim to enhance our understanding of drug characteristics,\nincluding their activity in biological contexts. However, a major challenge in\ndrug response (DR) prediction is model interpretability as it aids in the\nvalidation of findings. This is important in biomedicine, where models need to\nbe understandable in comparison with established knowledge of drug interactions\nwith proteins. drGAT, a graph deep learning model, leverages a heterogeneous\ngraph composed of relationships between proteins, cell lines, and drugs. drGAT\nis designed with two objectives: DR prediction as a binary sensitivity\nprediction and elucidation of drug mechanism from attention coefficients. drGAT\nhas demonstrated superior performance over existing models, achieving 78\\%\naccuracy (and precision), and 76\\% F1 score for 269 DNA-damaging compounds of\nthe NCI60 drug response dataset. To assess the model's interpretability, we\nconducted a review of drug-gene co-occurrences in Pubmed abstracts in\ncomparison to the top 5 genes with the highest attention coefficients for each\ndrug. We also examined whether known relationships were retained in the model\nby inspecting the neighborhoods of topoisomerase-related drugs. For example,\nour model retained TOP1 as a highly weighted predictive feature for irinotecan\nand topotecan, in addition to other genes that could potentially be regulators\nof the drugs. Our method can be used to accurately predict sensitivity to drugs\nand may be useful in the identification of biomarkers relating to the treatment\nof cancer patients.\n", " For generic binary forms $A_1,...,A_r$ of order $d$ we construct a class of\ncombinants $C = \\{\\C_q: 0 \\le q \\le r, q \\neq 1\\}$, to be called the Wronskian\ncombinants of the $A_i$. We show that the collection $C$ gives a projective\nimbedding of the Grassmannian $G(r,S_d)$, and as a corollary, any other\ncombinant admits a formula as an iterated transvectant in the $C$. Our second\nmain result characterizes those collections of binary forms which can arise as\nWronskian combinants. These collections are the ones such that an associated\nalgebraic differential equation has the maximal number of linearly independent\npolynomial solutions. Along the way we deduce some identities which connect\nWronskians with transvectants.\n", " We describe a method for bounding the set of exceptional integers not\nrepresented by a given additive form in terms of the exceptional set\ncorresponding to a subform. Illustrating our ideas with examples stemming from\nWaring's problem for cubes, we show, in particular, that the number of positive\nintegers not exceeding N, that fail to have a representation as the sum of six\ncubes of natural numbers, is O(N^{3/7}).\n", " We study the following inverse graph-theoretic problem: how many vertices\nshould a graph have given that it has a specified value of some parameter. We\nobtain asymptotic for the minimal number of vertices of the graph with the\ngiven number $n$ of maximal independent sets for a class of natural numbers\nthat can be represented as concatenation of periodic binary words.\n", " The present work demonstrates a robust protocol for probing localized\nelectronic structure in condensed-phase systems, operating in terms of a\nrecently proposed theory for decomposing the results of Kohn-Sham density\nfunctional theory in a basis of spatially localized molecular orbitals\n[Eriksen, J. Chem. Phys. 153, 214109 (2020)]. In an initial application to\nliquid, ambient water and the assessment of the solvation energy and the\nembedded dipole moment of H$_2$O in solution, we find that both properties are\namplified on average -- in accordance with expectation -- and that correlations\nare indeed observed to exist between them. However, the simulated\nsolvent-induced shift to the dipole moment of water is found to be\nsignificantly dampened with respect to typical literature values. The local\nnature of our methodology has further allowed us to evaluate the convergence of\nbulk properties with respect to the extent of the underlying one-electron basis\nset, ranging from single-$\\zeta$ to full (augmented) quadruple-$\\zeta$ quality.\nAlbeit a pilot example, our work paves the way towards future studies of local\neffects and defects in more complex phases, e.g., liquid mixtures and even\nsolid-state crystals.\n", " We explore the asymptotic distributions of sequences of integer-valued\nadditive functions defined on the symmetric group endowed with the Ewens\nprobability measure as the order of the group increases.\n Applying the method of factorial moments, we establish necessary and\nsufficient conditions for the weak convergence of distributions to discrete\nlaws.\n More attention is paid to the Poisson limit distribution. The particular case\nof the number-of-cycles with restricted lengths function is analyzed in more\ndetail. The results can be applied to statistics defined on random permutation\nmatrices.\n", " In this paper we use considerations of non-commutative geometry to deduce a\nmodel for QCD interactions. The model also explains within the same theoretical\nframework hitherto purely phenomenological characteristics of the quarks like\ntheir fractional charge, mass,handedness and confinement.\n", " CoV2019 has evolved to be much more dangerous than CoV2003. Experiments\nsuggest that structural rearrangements dramatically enhance CoV2019 activity.\nWe identify a new first stage of infection that precedes structural\nrearrangements by using biomolecular evolutionary theory to identify sequence\ndifferences enhancing viral attachment rates. We find a small cluster of four\nsingle mutations which show that CoV-2 has a new feature that promotes much\nstronger viral attachment and enhances contagiousness. The extremely dangerous\ndynamics of human coronavirus infection is a dramatic example of evolutionary\napproach of self-organized networks to criticality. It may favor a very\nsuccessful vaccine. The identified mutations can be used to test the present\ntheory experimentally. The theory also works well for the newer strains and\nexplains their increased contagiousness.\n", " Persistence diagrams have been widely used to quantify the underlying\nfeatures of filtered topological spaces in data visualization. In many\napplications, computing distances between diagrams is essential; however,\ncomputing these distances has been challenging due to the computational cost.\nIn this paper, we propose a persistence diagram hashing framework that learns a\nbinary code representation of persistence diagrams, which allows for fast\ncomputation of distances. This framework is built upon a generative adversarial\nnetwork (GAN) with a diagram distance loss function to steer the learning\nprocess. Instead of using standard representations, we hash diagrams into\nbinary codes, which have natural advantages in large-scale tasks. The training\nof this model is domain-oblivious in that it can be computed purely from\nsynthetic, randomly created diagrams. As a consequence, our proposed method is\ndirectly applicable to various datasets without the need for retraining the\nmodel. These binary codes, when compared using fast Hamming distance, better\nmaintain topological similarity properties between datasets than other\nvectorized representations. To evaluate this method, we apply our framework to\nthe problem of diagram clustering and we compare the quality and performance of\nour approach to the state-of-the-art. In addition, we show the scalability of\nour approach on a dataset with 10k persistence diagrams, which is not possible\nwith current techniques. Moreover, our experimental results demonstrate that\nour method is significantly faster with the potential of less memory usage,\nwhile retaining comparable or better quality comparisons.\n", " The methods of statistical physics of open systems are used for describing\nthe time dependence of economic characteristics (income, profit, cost, supply,\ncurrency etc.) and their correlations with each other. Nonlinear equations\n(analogies of known reaction-diffusion, kinetic, Langevin equation) describing\nappearance of bifurcations, self-sustained oscillational processes,\nself-organizations in economic phenomena are offered.\n", " While texture analysis is largely addressed for images, the comparison of the\ngeometric reliefs on surfaces embedded in the 3D space is still an open\nchallenge. Starting from the Local Binary Pattern (LBP) description originally\ndefined for images, we introduce the edge-Local Binary Pattern (edgeLBP) as a\nlocal description able to capture the evolution of repeated, geometric patterns\non surface meshes. Our extension is independent of the surface representation,\nindeed the edgeLBP is able to deal with surface tessellations characterized by\nnon-uniform vertex distributions and different types of faces, such as\ntriangles, quadrangles and, in general, convex polygons. Besides the desirable\nrobustness properties the edgeLBP exhibits over a number of examples, we show\nhow this description performs well for 3D pattern retrieval and compare our\nperformances with the participants to a recent 3D pattern retrieval and\nclassification contest.\n", " Using the ideas of E.I. Gordon we present and farther advance an approach,\nbased on nonstandard analysis, to simultaneous approximations of locally\ncompact abelian groups and their duals by (hyper)finite abelian groups, as well\nas to approximations of various types of Fourier transforms on them by the\ndiscrete Fourier transform. Combining some methods of nonstandard analysis and\nadditive combinatorics we prove the three Gordon's Conjectures which were open\nsince 1991 and are crucial both in the formulations and proofs of the LCA\ngroups and Fourier transform approximation theorems.\n", " Bacterial sensor systems can be used for the detection and measurement of\nmolecular signal concentrations. The dynamics of the sensor directly depend on\nthe biological properties of the bacterial sensor cells; manipulation of these\nfeatures in the wet lab enables the engineering and optimization of the\nbacterial sensor kinetics. This necessitates the development of biologically\nmeaningful computational models for bacterial sensors comprising a variety of\ndifferent molecular mechanisms, which further facilitates a systematic and\nquantitative evaluation of optimization strategies. In this work, we dissect\nthe detection chain of bacterial sensors from a mathematical perspective from\nwhich we derive, supported by wet-lab data, a complete computational model for\na Streptococcus mutans-based bacterial sensor as a case example. We address the\nengineering of bacterial sensors by investigating the impact of altered\nbacterial cell properties on the sensor response characteristics, specifically\nsensor sensitivity and response signal intensity. This is achieved through a\nsensitivity analysis targeting both the steady-state and transient sensor\nresponse characteristics. Alongside the demonstration of suitability of our\nmethodological approach, our analysis shows that an increase of sensor\nsensitivity, through a targeted manipulation of bacterial physiology, often\ncomes at the cost of generally diminished sensor response intensity.\n", " Huffman Codes are optimal Instantaneous Fixed-to-Variable (FV) codes in which\nevery source symbol can only be encoded by one codeword. Relaxing these\nconstraints permits constructing better FV codes. More specifically, recent\nwork has shown that AIFV-$m$ codes can beat Huffman coding. AIFV-$m$ codes\nconstruct am $m$-tuple of different coding trees between which the code\nalternates and are only almost instantaneous (AI). This means that decoding a\nword might require a delay of a finite number of bits.\n Current algorithms for constructing optimal AIFV-$m$ codes are iterative\nprocesses that construct progressively \"better sets\" of code trees. The\nprocesses have been proven to finitely converge to the optimal code but with no\nknown bounds on the convergence rate.\n This paper derives a geometric interpretation of the space of binary AIFV-$2$\ncodes, permitting the development of the first polynomially time-bounded\niterative procedures for constructing optimal AIFV codes.\n We first show that a simple binary search procedure can replace the current\niterative process to construct optimal binary AIFV-$2$ codes. We then describe\nhow to frame the problem as a linear programming with an exponential number of\nconstraints but a polynomial-time separability oracle. This permits using the\nGr\\\"otschel, Lov\\'asz and Schrijver ellipsoid method to solve the problem in a\npolynomial number of steps. While more complicated, this second method has the\npotential to lead to a polynomial time algorithm to construct optimal AIFV-$m$\ncodes for general $m$.\n", " A connected graph can be associated with two distinct evolution algebras. In\nthe first case, the structural matrix is the adjacency matrix of the graph\nitself. In the second case, the structural matrix is the transition\nprobabilities matrix of the symmetric random walk on the same graph. It is\nwell-known that, for a non-singular graph, both evolution algebras are\nisomorphic if, and only if, the graph is regular or biregular. Moreover,\nthrough examples and partial results, it has been conjectured that the same\nresult remains true for singular graphs. The purpose of this work is to provide\nnew examples supporting this conjecture.\n", " We give a proof of the existence of generalized definable locally compact\nmodels for arbitrary approximate subgroups via an application of topological\ndynamics in model theory. Our construction is simpler and shorter than the\noriginal one obtained by Hrushovski in ``Beyond the Lascar group'', and it uses\nonly basic model theory (mostly spaces of types and realizations of types). The\nmain tools are Ellis groups from topological dynamics considered for suitable\nspaces of types. However, we need to redevelop some basic theory of topological\ndynamics for suitable ``locally compact flows'' in place of (compact) flows. We\nalso prove that the generalized definable locally compact model which we\nconstructed is universal in an appropriate category. We note that the main\nresult yields structural information on definable generic subsets of definable\ngroups, with a more precise structural result for generics in the universal\ncover of $\\textrm{SL}_2(\\mathbb{R})$.\n", " Phase retrieval in real or complex Hilbert spaces is the task of recovering a\nvector, up to an overall unimodular multiplicative constant, from magnitudes of\nlinear measurements. In this paper, we assume that the vector is normalized,\nbut retain only qualitative, binary information about the measured magnitudes\nby comparing them with a threshold. In more specific, geometric terms, we\nchoose a sequence of subspaces in a real or complex Hilbert space and only\nrecord whether a given vector is closer to the subspace than to the\ncomplementary subspace. The subspaces have half the dimension of the Hilbert\nspace and are independent, uniformly distributed with respect to the action of\nthe orthogonal or unitary groups. The main goal of this paper is to find a\nfeasible algorithm for approximate recovery based on the information gained\nabout the vector from these binary questions and to establish error bounds for\nits approximate recovery. We provide a pointwise bound for fixed input vectors\nand a uniform bound that controls the worst-case scenario among all inputs.\nBoth bounds hold with high probability with respect to the choice of the\nsubspaces. For real or complex vectors of dimension $n$, the pointwise bound\nrequires $m \\ge C \\delta^{-2} n \\log(n)$ and the uniform bound $m \\ge C\n\\delta^{-2} n^2 \\log(\\delta^{-1} n)$ binary questions in order to achieve an\naccuracy of $\\delta$. The accuracy $\\delta$ is measured by the operator norm of\nthe difference between the rank-one orthogonal projections corresponding to the\nnormalized input vector and its approximate recovery.\n", " Persistent homology is a tool that can be employed to summarize the shape of\ndata by quantifying homological features. When the data is an object in\n$\\mathbb{R}^d$, the (augmented) persistent homology transform ((A)PHT) is a\nfamily of persistence diagrams, parameterized by directions in the ambient\nspace. A recent advance in understanding the PHT used the framework of\nreconstruction in order to find finite a set of directions to faithfully\nrepresent the shape, a result that is of both theoretical and practical\ninterest. In this paper, we improve upon this result and present an improved\nalgorithm for graph -- and, more generally one-skeleton -- reconstruction. The\nimprovement comes in reconstructing the edges, where we use a radial binary\n(multi-)search. The binary search employed takes advantage of the fact that the\nedges can be ordered radially with respect to a reference plane, a feature\nunique to graphs.\n", " Evidential reasoning is cast as the problem of simplifying the\nevidence-hypothesis relation and constructing combination formulas that possess\ncertain testable properties. Important classes of evidence as identifiers,\nannihilators, and idempotents and their roles in determining binary operations\non intervals of reals are discussed. The appropriate way of constructing\nformulas for combining evidence and their limitations, for instance, in\nrobustness, are presented.\n", " Networks of neurons in the brain encode preferred patterns of neural activity\nvia their synaptic connections. Despite receiving considerable attention, the\nprecise relationship between network connectivity and encoded patterns is still\npoorly understood. Here we consider this problem for networks of\nthreshold-linear neurons whose computational function is to learn and store a\nset of binary patterns (e.g., a neural code) as \"permitted sets\" of the\nnetwork. We introduce a simple Encoding Rule that selectively turns \"on\"\nsynapses between neurons that co-appear in one or more patterns. The rule uses\nsynapses that are binary, in the sense of having only two states (\"on\" or\n\"off\"), but also heterogeneous, with weights drawn from an underlying synaptic\nstrength matrix S. Our main results precisely describe the stored patterns that\nresult from the Encoding Rule -- including unintended \"spurious\" states -- and\ngive an explicit characterization of the dependence on S. In particular, we\nfind that binary patterns are successfully stored in these networks when the\nexcitatory connections between neurons are geometrically balanced -- i.e., they\nsatisfy a set of geometric constraints. Furthermore, we find that certain types\nof neural codes are \"natural\" in the context of these networks, meaning that\nthe full code can be accurately learned from a highly undersampled set of\npatterns. Interestingly, many commonly observed neural codes in cortical and\nhippocampal areas are natural in this sense. As an application, we construct\nnetworks that encode hippocampal place field codes nearly exactly, following\npresentation of only a small fraction of patterns. To obtain our results, we\nprove new theorems using classical ideas from convex and distance geometry,\nsuch as Cayley-Menger determinants, revealing a novel connection between these\nareas of mathematics and coding properties of neural networks.\n", " We introduce a tensor-based clustering method to extract sparse,\nlow-dimensional structure from high-dimensional, multi-indexed datasets. This\nframework is designed to enable detection of clusters of data in the presence\nof structural requirements which we encode as algebraic constraints in a linear\nprogram. Our clustering method is general and can be tailored to a variety of\napplications in science and industry. We illustrate our method on a collection\nof experiments measuring the response of genetically diverse breast cancer cell\nlines to an array of ligands. Each experiment consists of a cell line-ligand\ncombination, and contains time-course measurements of the early-signalling\nkinases MAPK and AKT at two different ligand dose levels. By imposing\nappropriate structural constraints and respecting the multi-indexed structure\nof the data, the analysis of clusters can be optimized for biological\ninterpretation and therapeutic understanding. We then perform a systematic,\nlarge-scale exploration of mechanistic models of MAPK-AKT crosstalk for each\ncluster. This analysis allows us to quantify the heterogeneity of breast cancer\ncell subtypes, and leads to hypotheses about the signalling mechanisms that\nmediate the response of the cell lines to ligands.\n", " A method of proving local continuity of concave functions on convex set\npossessing the $\\mu$-compactness property is presented. This method is based on\na special approximation of these functions.\n The class of $\\mu$-compact sets can be considered as a natural extension of\nthe class of compact metrizable subsets of locally convex spaces, to which\nparticular results well known for compact sets can be generalized.\n Applications of the obtained continuity conditions to analysis of different\nentropic characteristics of quantum systems and channels are considered.\n" ]

[ " We develop a thermodynamic theory in the non-equilibrium regime by extending\na theory based on the dimensionless (DL) minimum work principle previously\ndeveloped for a thermodynamic system-bath model [S.Koyanagi and\nY.Tanimura,J.Chem.Phys.160,(2024)]. Our results are described by\nnon-equilibrium thermodynamic potentials expressed in time-derivative form in\nterms of extensive and intensive variables. This is made possible through the\nincorporation of waste heat, which is equivalent to the loss work consumed by\nthe bath, into the definitions of thermodynamic potentials in a non-equilibrium\nregime. These potentials can be evaluated from the DL\nnon-equilibrium-to-equilibrium minimum work principle, which is derived from\nthe principle of DL minimum work and is equivalent to the second law of\nthermodynamics. We thus obtain the non-equilibrium Massieu-Planck potentials as\nentropic potentials and the non-equilibrium Helmholtz-Gibbs potentials as free\nenergies. Our results are verified numerically by simulating a Stirling engine\nconsisting of two isothermal and two thermostatic processes using the quantum\nFokker-Planck equations and the classical Kramers equation derived from the\nthermodynamic system-bath model. We then show that any thermodynamic process\ncan be analyzed using a non-equilibrium work diagram analogous to the\nequilibrium one for given time-dependent intensive variables. The results can\nbe used to develop efficient heat machines in non-equilibrium regimes.\n", " A generic and intuitive model for coherent energy transport in multiple\nminima systems coupled to a quantum mechanical bath is shown. Using a simple\nspin-boson system, we illustrate how a generic donor-acceptor system can be\nbrought into resonance using a narrow band of vibrational modes, such that the\ntransfer efficiency of an electron-hole pair (exciton) is made arbitrarily\nhigh. Coherent transport phenomena in nature are of renewed interest since the\ndiscovery that a photon captured by the light-harvesting complex (LHC) in\nphotosynthetic organisms can be conveyed to a chemical reaction centre with\nnear-perfect efficiency. Classical explanations of the transfer use stochastic\ndiffusion to model the hopping motion of a photo-excited exciton. This accounts\ninadequately for the speed and efficiency of the energy transfer measured in a\nseries of recent landmark experiments. Taking a quantum mechanical perspective\ncan help capture the salient features of the efficient part of that transfer.\nTo show the versatility of the model, we extend it to a multiple minima system\ncomprising seven-sites, reminiscent of the widely studied Fenna-Matthews-Olson\n(FMO) light-harvesting complex. We show that an idealised transport model for\nmultiple minima coupled to a narrow-band phonon can transport energy with\narbitrarily high efficiency.\n", " The Escherichia coli chemoreceptors form an extensive array that achieves\ncooperative and adaptive sensing of extracellular signals. The receptors\ncontrol the activity of histidine kinase CheA, which drives a non-equilibrium\nphosphorylation-dephosphorylation reaction cycle for response regulator CheY.\nRecent single-cell FRET measurements revealed that kinase activity of the array\nspontaneously switches between active and inactive states, with asymmetric\nswitching times that signify time-reversal symmetry breaking in the underlying\ndynamics. Here, we show that the asymmetric switching dynamics can be explained\nby a non-equilibrium lattice model, which considers both the dissipative\nreaction cycles of individual core units and the coupling between neighboring\nunits. The model reveals that large dissipation and near-critical coupling are\nrequired to explain the observed switching dynamics. Microscopically, the\nswitching time asymmetry originates from irreversible transition paths. The\nmodel shows that strong dissipation enables sensitive and rapid signaling\nresponse by relieving the speed-sensitivity trade-off, which can be tested by\nfuture single-cell experiments. Overall, our model provides a general framework\nfor studying biological complexes composed of coupled subunits that are\nindividually driven by dissipative cycles and the rich non-equilibrium physics\nwithin.\n", " Ultrafast non-thermal manipulation of magnetization by light relies on either\nindirect coupling of the electric field component of the light with spins via\nspin-orbit interaction or direct coupling between the magnetic field component\nand spins. Here we propose a novel scenario for coupling between the electric\nfield of light and spins via optical modification of the exchange interaction,\none of the strongest quantum effects, the strength of which can reach 1000\nTesla. We demonstrate that this isotropic opto-magnetic effect, which can be\ncalled the inverse magneto-refraction, is allowed in a material of any\nsymmetry. Its existence is corroborated by the experimental observation of THz\nemission by magnetic-dipole active spin resonances optically excited in a broad\nclass of iron oxides with a canted spin configuration. From its strength we\nestimate that a sub-picosecond laser pulse with a moderate fluence of ~ 1\nmJ/cm^2 acts as a pulsed effective magnetic field of 0.01 Tesla, arising from\nthe optically perturbed balance between the exchange parameters. Our findings\nare supported by a low-energy theory for the microscopic magnetic interactions\nbetween non-equilibrium electrons subjected to an optical field which suggests\na possibility to modify the exchange interactions by light over 1 %.\n", " We show that the gauge symmetry of the nonequilibrium quantum transport of\nChern insulator in a uniform electric field is governed by the Wilson line of\nparallel transport operator coupled with the dynamical translation operator.\nThis is dictated by the minimal coupling of derivatives with gauge fields in U\n(1) gauge theory. This parallel transport symmetry consideration leads to the\ninteger quantum Hall effect in electrical conductivity obtained to first-order\ngradient expansion of the nonequilibrium quantum transport equations.\n", " Electron scattering problem in the monolayer graphene with short-range\nimpurities is considered. The main novel element in the suggested model is the\nband asymmetry of the defect potential in the 2+1-dimensional Dirac equation.\nThis asymmetry appears naturally if the defect violates the symmetry between\nsublattices. Our goal in the present paper is to take into account a local band\nasymmetry violation arising due to the defect presence. We analyze the effect\nof the electron scattering on the electronic transport parameters in the\nmonolayer graphene. The explicit exact formula obtained for S-matrix for the\nsuggested delta-shell potential model allowed us to study the asymptotic\nbehavior of such characteristics as scattering phases, transport cross section,\nthe transport relaxation time and the conductivity for small values of the\nFermi energy. The obtained results are in a good agreement with the\nexperimental data which shows that the considered model is reasonable.\n", " We present a full theoretical and experimental study of the dynamics and\nenergy distribution of non-equilibrium quasiparticles in superconducting tunnel\njunctions (STJs). STJs are often used for single-photon spectrometers, where\nthe numbers of quasiparticles excited by a photon provide a measure of the\nphoton energy. The magnitude and fluctuations of the signal current in STJ\ndetectors are in large part determined by the quasiparticle dynamics and energy\ndistribution during the detection process. We use this as motivation to study\nthe transport and energy distribution of non-equilibrium quasiparticles excited\nby x-ray photons in a lateral, imaging junction configuration. We present a\nfull numerical model for the tunneling current of the major physical processes\nwhich determine the signal. We find that a diffusion framework models the\nquasiparticle dynamics well and that excited quasiparticles do not equilibrate\nto the lattice temperature during the timescales for tunneling. We extract\nphysical timescales from the measured data, make comparisons with existing\ntheories, and comment on implications for superconducting mesoscopic systems\nand single-photon detectors.\n", " Recently, transition metal dichalcogenides (TMDCs) semiconductors have been\nutilized for investigating quantum phenomena because of their unique band\nstructures and novel electronic properties. In a quantum dot (QD), electrons\nare confined in all lateral dimensions, offering the possibility for detailed\ninvestigation and controlled manipulation of individual quantum systems. Beyond\nthe definition of graphene QDs by opening an energy gap in nanoconstrictions,\nwith the presence of a bandgap, gate-defined QDs can be achieved on TMDCs\nsemiconductors. In this paper, we review the confinement and transport of QDs\nin TMDCs nanostructures. The fabrication techniques for demonstrating\ntwo-dimensional (2D) materials nanostructures such as field-effect transistors\nand QDs, mainly based on e-beam lithography and transfer assembly techniques\nare discussed. Subsequently, we focus on transport through TMDCs nanostructures\nand QDs. With steady improvement in nanoscale materials characterization and\nusing graphene as a springboard, 2D materials offer a platform that allows\ncreation of heterostructure QDs integrated with a variety of crystals, each of\nwhich has entirely unique physical properties.\n", " Topological phases of matter have recently attracted a rather notable\nattention in the community dealing with the holographic methods applied to\nstrongly interacting condensed matter systems. In particular, holographic\nmodels for gapless Weyl and multi-Weyl semimetals, characterized on a lattice\nby the monopole-antimonopole defects of the Berry curvature in momentum space,\nwere recently formulated. In this paper, motivated by the quest for finding\ntopological holographic phases, we show that holographic model for multi-Weyl\nsemimetals features a rather rich landscape of phases. In particular, it\nincludes a novel phase which we dub $xy$ nematic, stable at strong coupling, as\nwe explicitly show by the free energy and the quasi-normal mode analyses.\nFurthermore, we provide its characterization through the anomalous transport\ncoefficients. We hope that our findings will motivate future works exploring\nthe holographic realizations of the topological phases.\n", " Repeated quantum measurements can generate effective new non-equilibrium\ndynamics in matter. Here we combine such a measurement driven system with\ndisorder. In particular, we investigate the diffusive behavior in the system\nand the effect of various types of disorder on the measurement induced chiral\ntransport protocol [1]. We begin by characterizing the diffusive behavior\nproduced by the measurements themselves in a clean system. We then examine the\nedge flow of particles per measurement cycle for three different types of\ndisorder: site dilution, lattice distortion, and disorder in onsite chemical\npotential. In the quantum Zeno limit, the effective descriptions for the\ndisordered measurement system with lattice distortions and random onsite\npotential can be modelled as a classical stochastic model, and the overall\neffect of increasing these disorders induces a crossover from perfect flow to\nzero transport. On the other hand if vacancies are present in the lattice the\nflow of particles per measurement cycle undergoes a percolation phase\ntransition from unity to zero with percolation threshold $p_c \\approx 0.26$,\nwith critical exponent $\\nu \\approx 1.35$. We also present numerical results\naway from Zeno limit and note that the overall effect of moving away from the\nZeno effect is to reduce particle flow per cycle when the measurement frequency\nin our protocol is reduced.\n", " The multilevel blocking algorithm recently proposed as a possible solution to\nthe sign problem in path-integral Monte Carlo simulations has been extended to\nsystems with long-ranged interactions along the Trotter direction. As an\napplication, new results for the real-time quantum dynamics of the spin-boson\nmodel are presented.\n", " In this paper we propose and study a novel optimal transport based\nregularization of linear dynamic inverse problems. The considered inverse\nproblems aim at recovering a measure valued curve and are dynamic in the sense\nthat (i) the measured data takes values in a time dependent family of Hilbert\nspaces, and (ii) the forward operators are time dependent and map, for each\ntime, Radon measures into the corresponding data space. The variational\nregularization we propose is based on dynamic (un-)balanced optimal transport\nwhich means that the measure valued curves to recover (i) satisfy the\ncontinuity equation, i.e., the Radon measure at time $t$ is advected by a\nvelocity field $v$ and varies with a growth rate $g$, and (ii) are penalized\nwith the kinetic energy induced by $v$ and a growth energy induced by $g$. We\nestablish a functional-analytic framework for these regularized inverse\nproblems, prove that minimizers exist and are unique in some cases, and study\nregularization properties. This framework is applied to dynamic image\nreconstruction in undersampled magnetic resonance imaging (MRI), modelling\nrelevant examples of time varying acquisition strategies, as well as patient\nmotion and presence of contrast agents.\n", " Jets and photons could play an important role in finding the transport\ncoefficients of the quark-gluon plasma. To this end we analyze their\ninteraction with a non-equilibrium quark-gluon plasma. Using new\nfield-theoretical tools we derive two-point correlators for the plasma which\nshow how instabilities evolve in time. This allows us, for the first time, to\nderive finite rates of interaction with the medium. We furthermore show that\ncoherent, long-wavelength instability fields in the Abelian limit do not modify\nthe rate of photon emission or jet-medium interaction.\n", " Quantum dot (QD) solids are being widely exploited as a solution-processable\ntechnology to develop photovoltaic, light-emission, and photo-detection\ndevices. Charge transport in these materials is the result of a compromise\nbetween confinement at the individual QD level and electronic coupling among\nthe different nanocrystals in the ensemble. While this is commonly achieved by\nligand engineering in colloidal-based systems, ligand-free QD assemblies have\nrecently emerged as an exciting alternative where nanostructures can be\ndirectly grown into porous matrices with optical quality as well as control\nover their connectivity and hence charge transport properties. In this context,\nwe present a complete photophysical study comprising fluence and\ntemperature-dependent time-resolved spectroscopy to study carrier dynamics in\nligand-free QD networks with gradually varying degrees of interconnectivity,\nwhich we achieve by changing the average distance between the QDs. Analysis of\nthe photoluminescence and absorption properties of the QD assemblies, involving\nboth static and time-resolved measurements, allows us to identify the weight of\nthe different recombination mechanisms, both radiative and non-radiative, as a\nfunction of QD connectivity. We propose a picture where carrier diffusion,\nwhich is needed for any optoelectronic application and implies inter-particle\ntransport, gives rise to the exposure of carriers to a larger defect landscape\nthan in the case of isolated QDs. The use of a broad range of fluences permits\nextracting valuable information for applications demanding either low or high\ncarrier injection levels and highlighting the relevance of a judicious design\nto balance recombination and diffusion.\n", " We investigate from first-principles the change in transport properties of a\ntwo-dimensional azobenzene monolayer sandwiched between two Au electrodes that\nundergoes molecular switching. We focus on transport differences between a\nchemisorbed and physisorbed top monolayer-electrode contact. The conductance of\nthe monolayer junction with a chemisorbed top contact is higher in trans\nconfiguration, in agreement with the previous theoretical predictions of\none-dimensional single molecule junctions. However, with a physisorbed top\ncontact, the \"ON\" state with larger conductance is associated with the cis\nconfiguration due to a reduced effective tunneling pathway by switching from\ntrans to cis, which successfully explains recently experimental measurements of\nazobenzene monolayer junctions. A simple model is developed to explain electron\ntransmission across subsystems in the molecular junction. We also discuss the\neffects of monolayer packing density, molecule tilt angle, and contact geometry\non the calculated transmission functions. In particular, we find that a\ntip-like contact with chemisorption significantly affects the electric current\nthrough the cis monolayer, leading to highly asymmetric current-voltage\ncharacteristics as well as large negative differential resistance behavior.\n", " Focus of the chapter is on the theoretical approaches aimed to analyze\nthermoelectric properties at the nanoscale. We discuss several relevant\ntheoretical approaches for different set-ups of nano-devices providing\nestimations of the thermoelectric parameters in the linear and non-linear\nregime, in particular the thermoelectric figure of merit and the\npower-efficiency trade-off. Moreover, we analyze the role of not only\nelectronic, but also of vibrational degrees of freedom. First, nanoscale\nthermoelectric phenomena are considered in the quantum coherent regime using\nthe Landauer-B\\\"uttiker method and focusing on effects of energy filtering.\nThen, we analyze the effects of many-body couplings between nano-structure\ndegrees of freedom, such as electron-electron and electron-vibration\ninteractions, which can strongly affect the thermoelectric conversion. In\nparticular, we discuss the enhancement of the thermoelectric figure of merit in\nthe Coulomb blockade regime for a quantum dot model starting from the master\nequation for charge state probabilities and the tunneling rates through the\nelectrodes. Finally, within the non-equilibrium Green function formalism, we\nquantify the reduction of the thermoelectric performance in simple models of\nmolecular junctions due to the effects of the electron-vibration coupling and\nphonon transport at room temperature.\n", " A Z3 symmetric generalization of the Dirac equation was proposed in recent\nseries of papers, where its properties and solutions discussed. The generalized\nDirac operator acts on \"coloured spinors\" composed out of six Pauli spinors,\ndescribing three colours and particle-antiparticle degrees of freedom\ncharacterizing a single quark state, thus combining Z2 x Z_2 x Z_3 symmetries\nof 12-component generalized wave functions. Spinorial representation of the\nZ3-graded generalized Lorentz algebra was introduced, leading to the appearance\nof extra Z2 x Z2 x Z3 symmetries, probably englobing the symmetries of isospin,\nflavors and families. The present article proposes a construction of Z3-graded\nextension of the Poincar\\'e algebra. It turns out that such a generalization\nrequires introduction of extended 12-dimensional Minkowskian space-time\ncontaining the usual 4-dimensional space-time as a subspace, and two other\nmutually conjugate \"replicas\" with complex-valued vectors and metric tensors.\nRepresentation in terms of differential operators and generalized Casimir\noperators are introduced and their symmetry properties are briefly discussed.\n", " The present study is elaborated to investigate the validity of\nthermodynamical laws in a modified teleparallel gravity based on higher-order\nderivatives terms of torsion scalar. For this purpose, we consider spatially\nflat FRW model filled with perfect fluid matter contents. Firstly, we explore\nthe possibility of existence of equilibrium as well as non-equilibrium picture\nof thermodynamics in this extended version of teleprallel gravity. Here, we\npresent the first law and the generalized second law of thermodynamics (GSLT)\nusing Hubble horizon. It is found that non-equilibrium description of\nthermodynamics exists in this theory with the presence of an extra term called\nas entropy production term. We also establish GSLT using the logarithmic\ncorrected entropy. Further, by taking the equilibrium picture, we discuss\nvalidity of GSLT at Hubble horizon for two $F$ different models. Using Gibbs\nlaw and the assumption that temperature of matter within Hubble horizon is\nsimilar to itself, we use different cases for choices scale factors to discuss\nthe GSLT validity graphically in all scenarios. It is found that the GSLT is\nsatisfied for a specified range of free parameters in all cases.\n", " There must exist a reformulation of quantum field theory, even at low\nenergies, which does not depend on classical time. The octonionic theory\nproposes such a reformulation, leading to a pre-quantum pre-spacetime theory.\nThe ingredients for constructing such a theory, which is also a unification of\nthe standard model with gravitation, are : (i) the pre-quantum theory of trace\ndynamics: a matrix-valued Lagrangian dynamics, (ii) the spectral action\nprinciple of non-commutative geometry, (iii) the number system known as the\noctonions, for constructing a non-commutative manifold and for defining\nelementary particles via Clifford algebras, (iv) a Lagrangian with $E_8 \\times\nE_8$ symmetry. The split bioctonions define a sixteen dimensional space (with\nleft-right symmetry) whose geometry (evolving in Connes time) relates to the\nfour known fundamental forces, while predicting two new forces, $SU(3)_{grav}$\nand $U(1)_{grav}$. This latter interaction is possibly the theoretical origin\nof MOND. Coupling constants of the standard model result from left-right\nsymmetry breaking, and their values are theoretically determined by the\ncharacteristic equation of the exceptional Jordan algebra of the octonions. The\nquantum-to-classical transition, precipitated by the entanglement of a critical\nnumber of fermions, is responsible for the emergence of classical spacetime,\nand also for the familiar formulation of quantum theory on a spacetime\nbackground.\n", " A generalized continuity equation extending the ordinary continuity equation\nhas been found using quanternions. It is shown to be compatible with Dirac,\nSchrodinger, Klein-Gordon and diffusion equations. This generalized equation is\nLorentz invariant. The transport properties of electrons are found to be\ngoverned by Schrodinger-like equation and not by the diffusion equation.\n", " These notes constitute a sort of Crash Course in Optimal Transport Theory.\nThe different features of the problem of Monge-Kantorovitch are treated,\nstarting from convex duality issues. The main properties of space of\nprobability measures endowed with the distances $W_p$ induced by optimal\ntransport are detailed. The key tools to put in relation optimal transport and\nPDEs are provided.\n", " We provide a broad review of fundamental electronic properties of\ntwo-dimensional graphene with the emphasis on density and temperature dependent\ncarrier transport in doped or gated graphene structures. A salient feature of\nour review is a critical comparison between carrier transport in graphene and\nin two-dimensional semiconductor systems (e.g. heterostructures, quantum wells,\ninversion layers) so that the unique features of graphene electronic properties\narising from its gap- less, massless, chiral Dirac spectrum are highlighted.\nExperiment and theory as well as quantum and semi-classical transport are\ndiscussed in a synergistic manner in order to provide a unified and\ncomprehensive perspective. Although the emphasis of the review is on those\naspects of graphene transport where reasonable consensus exists in the\nliterature, open questions are discussed as well. Various physical mechanisms\ncontrolling transport are described in depth including long- range charged\nimpurity scattering, screening, short-range defect scattering, phonon\nscattering, many-body effects, Klein tunneling, minimum conductivity at the\nDirac point, electron-hole puddle formation, p-n junctions, localization,\npercolation, quantum-classical crossover, midgap states, quantum Hall effects,\nand other phenomena.\n", " QCD in non-integer $d=4-2\\epsilon$ space-time dimensions enjoys conformal\ninvariance at the special fine-tuned value of the coupling. Counterterms for\ncomposite operators in minimal subtraction schemes do not depend on $\\epsilon$\nby construction, and therefore the renormalization group equations for\ncomposite operators in physical (integer) dimensions inherit conformal\nsymmetry. This observation can be used to restore the complete evolution\nkernels that take into account mixing with the operators containing total\nderivatives from their eigenvalues (anomalous dimensions). Using this approach\nwe calculate the two-loop (NLO) evolution kernels for the leading twist\nflavor-singlet operators in the position space (light-ray operator)\nrepresentation. As the main result of phenomenological relevance, in this way\nwe are able to confirm the evolution equations of flavor-singlet generalized\nhadron parton distributions derived earlier by Belitsky and M\\\"uller using a\ndifferent approach.\n", " In our present investigation, we are concerned with the Kantorovich variant\nof Lupa\\c{s}-Stancu operators based on Polya distribution with Pochhammer\n$k$-symbol. We briefly give some basic properties of the generalized operators\nand by making use of these results, we investigate convergence properties of\nthe studied operators. Furthermore, the rate of convergence of these operators\nis obtained and Voronovskaja type theorem for the pointwise approximation is\nestablished. Then we construct bivariate generalization of the operators and we\ndiscuss some convergence properties. Finally, taking into account some\nillustrative graphics, we conclude our study with the comparison of the rate of\nconvergence between our operators and other operators which are mentioned in\nthe paper.\n", " We consider a linear size-structured population model with diffusion in the\nsize-space. Individuals are recruited into the population at arbitrary sizes.\nThe model is equipped with generalized Wentzell-Robin (or dynamic) boundary\nconditions. This allows modelling of \"adhesion\" at extremely small or large\nsizes. We establish existence and positivity of solutions by showing that\nsolutions are governed by a positive quasicontractive semigroup of linear\noperators on the biologically relevant state space. This is carried out via\nestablishing dissipativity of a suitably perturbed semigroup generator. We also\nshow that solutions of the model exhibit balanced exponential growth, that is\nour model admits a finite dimensional global attractor. In case of strictly\npositive fertility we are able to establish that solutions in fact exhibit\nasynchronous exponential growth.\n", " Recently, utilizing reinforcement learning (RL) to generate molecules with\ndesired properties has been highlighted as a promising strategy for drug\ndesign. A molecular docking program - a physical simulation that estimates\nprotein-small molecule binding affinity - can be an ideal reward scoring\nfunction for RL, as it is a straightforward proxy of the therapeutic potential.\nStill, two imminent challenges exist for this task. First, the models often\nfail to generate chemically realistic and pharmacochemically acceptable\nmolecules. Second, the docking score optimization is a difficult exploration\nproblem that involves many local optima and less smooth surfaces with respect\nto molecular structure. To tackle these challenges, we propose a novel RL\nframework that generates pharmacochemically acceptable molecules with large\ndocking scores. Our method - Fragment-based generative RL with Explorative\nExperience replay for Drug design (FREED) - constrains the generated molecules\nto a realistic and qualified chemical space and effectively explores the space\nto find drugs by coupling our fragment-based generation method and a novel\nerror-prioritized experience replay (PER). We also show that our model performs\nwell on both de novo and scaffold-based schemes. Our model produces molecules\nof higher quality compared to existing methods while achieving state-of-the-art\nperformance on two of three targets in terms of the docking scores of the\ngenerated molecules. We further show with ablation studies that our method,\npredictive error-PER (FREED(PE)), significantly improves the model performance.\n", " We find that studying the simplest of the coupled non-equilibrium growth\nequations of Barabasi by self-consistent mode coupling requires the use of\ndressed vertices. Using the vertex renormalization, we find a roughness\nexponent which already in the leading order is quite close to the numerical\nvalue.\n", " The temperature dependence of electric transport properties of single-layer\nand few-layer graphene at large charge doping is of great interest both for the\nstudy of the scattering processes dominating the conductivity at different\ntemperatures and in view of the theoretically predicted possibility to reach\nthe superconducting state in such extreme conditions. Here we present the\nresults obtained in 3-, 4- and 5-layer graphene devices down to 3.5 K, where a\nlarge surface charge density up to about 6.8x10^14 cm^(-2) has been reached by\nemploying a novel polymer electrolyte solution for the electrochemical gating.\nIn contrast with recent results obtained in single-layer graphene, the\ntemperature dependence of the sheet resistance between 20 K and 280 K shows a\nlow-temperature dominance of a T^2 component - that can be associated with\nelectron-electron scattering - and, at about 100 K, a crossover to the classic\nelectron-phonon regime. Unexpectedly this crossover does not show any\ndependence on the induced charge density, i.e. on the large tuning of the Fermi\nenergy.\n", " Hypergraphs capture multi-way relationships in data, and they have\nconsequently seen a number of applications in higher-order network analysis,\ncomputer vision, geometry processing, and machine learning. In this paper, we\ndevelop theoretical foundations for studying the space of hypergraphs using\ningredients from optimal transport. By enriching a hypergraph with probability\nmeasures on its nodes and hyperedges, as well as relational information\ncapturing local and global structures, we obtain a general and robust framework\nfor studying the collection of all hypergraphs. First, we introduce a\nhypergraph distance based on the co-optimal transport framework of Redko et al.\nand study its theoretical properties. Second, we formalize common methods for\ntransforming a hypergraph into a graph as maps between the space of hypergraphs\nand the space of graphs, and study their functorial properties and Lipschitz\nbounds. Finally, we demonstrate the versatility of our Hypergraph Co-Optimal\nTransport (HyperCOT) framework through various examples.\n", " Next-nearest-neighbor Heisenberg chain plays important roles in solid state\nphysics, such as predicting exotic electric properties of two-dimensional\nmaterials or magnetic properties of organic compounds. Direct experimental\nstudies of the many-body electron systems or spin systems associating to these\nmaterials are challenging tasks, while optical simulation provides an effective\nand economical way for immediate observation. Comparing with bulk optics,\nintegrated optics are more of fascinating for steady, large scale and long-time\nevolution simulations. Photonic crystal is an artificial microstructure\nmaterial with multiple methods to tune the propagation properties, which are\nessential for various simulation tasks. Here we report for the first time an\nexperimental simulation of next-nearest-neighbor Heisenberg chain with an\nintegrated optical chip of photonic crystal waveguide array. The use of\nphotonic crystal enhances evanescent field thus allows coupling between\nnext-nearest-neighbor waveguides in such a planar waveguide array, without\nbreaking the weak coupling condition of the coupled mode equation.\nParticularly, similarities between the model and coherent light propagation\ncould reach 0.99 in numerical simulations and 0.89 in experiment. Localization\neffect induced by second-order coupling and coupling strengthening with\nincreasing wavelengths were also revealed in both numerical simulations and\nexperiments. The platform proposed here is compatible with mature complementary\nmetal oxide semiconductor technology thus possesses the potential for\nlarger-scale problems and photonic crystals further allows simulations of\nspecific target Hamiltonians.\n", " A cross-shaped diffusive system with two superconducting and two normal\nelectrodes is considered. A voltage $eV < \\Delta$ is applied between the normal\nleads. Even in the absence of average current through the superconducting\nelectrodes their presence increases the shot noise at the normal electrodes and\ndoubles it in the case of a strong coupling to the superconductors. The\nnonequilibrium noise at the superconducting electrodes remains finite even in\nthe case of a vanishingly small transport current due to the absence of energy\ntransfer into the superconductors. This noise is suppressed by\nelectron-electron scattering at sufficiently high voltages.\n", " Electron transport in realistic physical and chemical systems often involves\nthe non-trivial exchange of energy with a large environment, requiring the\ndefinition and treatment of open quantum systems. Because the time evolution of\nan open quantum system employs a non-unitary operator, the simulation of open\nquantum systems presents a challenge for universal quantum computers\nconstructed from only unitary operators or gates. Here we present a general\nalgorithm for implementing the action of any non-unitary operator on an\narbitrary state on a quantum device. We show that any quantum operator can be\nexactly decomposed as a linear combination of at most four unitary operators.\nWe demonstrate this method on a two-level system in both zero and finite\ntemperature amplitude damping channels. The results are in agreement with\nclassical calculations, showing promise in simulating non-unitary operations on\nintermediate-term and future quantum devices.\n", " Mixing (or quasirandom) properties of the natural transition matrix\nassociated to a graph can be quantified by its distance to the complete graph.\nDifferent mixing properties correspond to different norms to measure this\ndistance. For dense graphs, two such properties known as spectral expansion and\nuniformity were shown to be equivalent in seminal 1989 work of Chung, Graham\nand Wilson. Recently, Conlon and Zhao extended this equivalence to the case of\nsparse vertex transitive graphs using the famous Grothendieck inequality. Here\nwe generalize these results to the non-commutative, or `quantum', case, where a\ntransition matrix becomes a quantum channel. In particular, we show that for\nirreducibly covariant quantum channels, expansion is equivalent to a natural\nanalog of uniformity for graphs, generalizing the result of Conlon and Zhao.\nMoreover, we show that in these results, the non-commutative and commutative\n(resp.) Grothendieck inequalities yield the best-possible constants.\n", " We examine the role of consistency with causality and quantum mechanics in\ndetermining the properties of gravitation. We begin by examining two different\nclasses of interacting theories of massless spin 2 particles -- gravitons. One\ninvolves coupling the graviton with the lowest number of derivatives to matter,\nthe other involves coupling the graviton with higher derivatives to matter,\nmaking use of the linearized Riemann tensor. The first class requires an\ninfinite tower of terms for consistency, which is known to lead uniquely to\ngeneral relativity. The second class only requires a finite number of terms for\nconsistency, which appears as another class of theories of massless spin 2. We\nrecap the causal consistency of general relativity and show how this fails in\nthe second class for the special case of coupling to photons, exploiting\nrelated calculations in the literature. In a companion paper [1] this result is\ngeneralized to a much broader set of theories. Then, as a causal modification\nof general relativity, we add light scalar particles and recap the generic\nviolation of universal free-fall they introduce and its quantum resolution.\nThis leads to a discussion of a special type of scalar-tensor theory; the\n$F(\\mathcal{R})$ models. We show that, unlike general relativity, these models\ndo not possess the requisite counterterms to be consistent quantum effective\nfield theories. Together this helps to remove some of the central assumptions\nmade in deriving general relativity.\n", " We propose and study a class-expansion/innovation/loss model of genome\nevolution taking into account biological roles of genes and their constituent\ndomains. In our model numbers of genes in different functional categories are\ncoupled to each other. For example, an increase in the number of metabolic\nenzymes in a genome is usually accompanied by addition of new transcription\nfactors regulating these enzymes. Such coupling can be thought of as a\nproportional \"recipe\" for genome composition of the type \"a spoonful of sugar\nfor each egg yolk\". The model jointly reproduces two known empirical laws: the\ndistribution of family sizes and the nonlinear scaling of the number of genes\nin certain functional categories (e.g. transcription factors) with genome size.\nIn addition, it allows us to derive a novel relation between the exponents\ncharacterising these two scaling laws, establishing a direct quantitative\nconnection between evolutionary and functional categories. It predicts that\nfunctional categories that grow faster-than-linearly with genome size to be\ncharacterised by flatter-than-average family size distributions. This relation\nis confirmed by our bioinformatics analysis of prokaryotic genomes. This proves\nthat the joint quantitative trends of functional and evolutionary classes can\nbe understood in terms of evolutionary growth with proportional recipes.\n", " The recent discovery of methods to isolate graphene, a one-atom-thick layer\nof crystalline carbon, has raised the possibility of a new class of\nnano-electronics devices based on the extraordinary electrical transport and\nunusual physical properties of this material. However, the experimental\nrealization of devices displaying these properties was, until now, hampered by\nthe influence of the ambient environment, primarily the substrate. Here we\nreport on the fabrication of Suspended Graphene devices and on studies of their\nelectrical transport properties. In these devices, environmental disturbances\nwere minimized allowing unprecedented access to the intrinsic properties of\ngraphene close to the Dirac Point (DP) where the energy dispersion of the\ncarriers and their density-of-states vanish linearly giving rise to a range of\nexotic physical properties. We show that charge inhomogeneity is reduced by\nalmost one order of magnitude compared to that in Non-Suspended Graphene\ndevices. Moreover, near the DP, the mobility exceeds 100,000 cm2/Vs,\napproaching theoretical predictions for evanescent transport in the ballistic\nmodel.\n", " We study the equilibrium instability of chiral quarkonia in a plasma in the\npresence of constant magnetic and electric field and at finite axial chemical\npotential using AdS/CFT duality. The model in use is a supersymmetric QCD at\nlarge 't$\\,$Hooft coupling and number of colors. We show that the presence of\nthe magnetic field and the axial chemical potential even in the absence of the\nelectric field make the system unstable. In a gapped system, a stable/unstable\nequilibrium state phase transition is observed and the initial transition\namplitude of the equilibrium state to the non-equilibrium state is\ninvestigated. We demonstrate that at zero temperature and large magnetic field\nthe instability grows linearly by increasing the quarkonium binding energy. In\nthe constant electric and magnetic field, the system is in a equilibrium state\nif the Ohm's law and the chiral magnetic effect cancel their effects. This\nhappens in a sub-space of $(E,B,T,\\mu_5)$ space with constraint equation\n$\\sigma_B B =- \\sigma E$, where $\\sigma$ and $\\sigma_B$ are called electric and\nchiral magnetic conductivity, respectively. We analyze the decay rate of a\ngapless system when this constraint is slightly violated.\n", " Boolean, free and monotone cumulants as well as relations among them, have\nproven to be important in the study of non-commutative probability theory.\nQuite notably, Boolean cumulants were successfully used to study free infinite\ndivisibility via the Boolean Bercovici--Pata bijection. On the other hand, in\nrecent years the concept of infinitesimal non-commutative probability has been\ndeveloped, together with the notion of infinitesimal cumulants which can be\nuseful in the context of combinatorial questions.\n In this paper, we show that the known relations among free, Boolean and\nmonotone cumulants still hold in the infinitesimal framework. Our approach is\nbased on the use of Grassmann algebra. Formulas involving infinitesimal\ncumulants can be obtained by applying a formal derivation to known formulas.\n The relations between the various types of cumulants turn out to be captured\nvia the shuffle algebra approach to moment-cumulant relations in\nnon-commutative probability theory. In this formulation, (free, Boolean and\nmonotone) cumulants are represented as elements of the Lie algebra of\ninfinitesimal characters over a particular combinatorial Hopf algebra. The\nlatter consists of the graded connected double tensor algebra defined over a\nnon-commutative probability space and is neither commutative nor cocommutative.\nIn this note it is shown how the shuffle algebra approach naturally extends to\nthe notion of infinitesimal non-commutative probability space. The basic step\nconsists in replacing the base field as target space of linear Hopf algebra\nmaps by the Grassmann algebra over the base field. We also consider the\ninfinitesimal analog of the Boolean Bercovici--Pata map.\n", " The \"textbook\" phonon mean free path (MFP) of heat carrying phonons in\nsilicon at room temperature is ~40 nm. However, a large contribution to the\nthermal conductivity comes from low-frequency phonons with much longer MFPs. We\npresent a simple experiment demonstrating that room temperature thermal\ntransport in Si significantly deviates from the diffusion model already at\nmicron distances. Absorption of crossed laser pulses in a freestanding silicon\nmembrane sets up a sinusoidal temperature profile that is monitored via\ndiffraction of a probe laser beam. By changing the period of the thermal\ngrating we vary the heat transport distance within the range ~1-10 {\\mu}m. At\nsmall distances, we observe a reduction in the effective thermal conductivity\nindicating a transition from the diffusive to the ballistic transport regime\nfor the low-frequency part of the phonon spectrum.\n", " We report experimental observation of electrically-tunable coherent perfect\nabsorption (CPA) of terahertz (THz) radiation in graphene. We develop a\nreflection-type tunable THz cavity formed by a large-area graphene layer, a\nmetallic reflective electrode and an electrolytic medium in between. Ionic\ngating in the THz cavity allows us to tune the Fermi energy of graphene up to\n1eV and to achieve critical coupling condition at 2.8 THz with absorption of\n99%. With the enhanced THz absorption, we were able to measure the Fermi energy\ndependence of the transport scattering time of highly doped graphene.\nFurthermore, we demonstrate flexible active THz surfaces that yield large\nmodulation in the THz reflectivity with low insertion losses. We anticipate\nthat the gate-tunable CPA will lead efficient active THz optoelectronics\napplications.\n", " We show that effective coupling strengths between ungapped and gapped quarks\nin the two-flavor color superconducting (2SC) phase are renormalized by\nlogarithmic quantum corrections. We obtain a set of coupled\nrenormalization-group (RG) equations for two distinct effective coupling\nstrengths arising from gluon exchanges carrying different color charges. The\ndiagram of RG flow suggests that both of the coupling strengths evolve into a\nstrong-coupling regime as we decrease the energy scale toward the Fermi\nsurface. This is a characteristic behavior observed in the Kondo effect, which\nhas been known to occur in the presence of impurity scatterings via non-Abelian\ninteractions. We propose a novel Kondo effect emerging without doped\nimpurities, but with the gapped quasiexcitations and the residual SU(2) color\nsubgroup intrinsic in the 2SC phase, which we call the 2SC Kondo effect.\n", " Anharmonicities provide a wealth of information about the vibrational\ndynamics, mode coupling and energy transfer within a polyatomic system. In this\ncontribution we show how driven molecular dynamics trajectories can be used to\nextract anharmonicity properties under very short times of a few hundreds of\nvibrational periods, using two exciting fields at- and slightly off-resonance.\nDetailed analyses on generic quartic potential energy surfaces and applications\nto various model systems are presented, giving good agreement with perturbation\ntheory. Application to a realistic molecule, cubane (C$_8$H$_8$), modeled with\na tight-binding quantum force field, further indicates how the method can be\napplied in practical cases.\n", " Querying new information from knowledge sources, in general, and published\nliterature, in particular, aims to provide precise and quick answers to\nquestions raised about a system under study. In this paper, we present\nACCORDION (Automated Clustering Conditional On Relating Data of Interactions tO\na Network), a novel tool and a methodology to enable efficient answering of\nbiological questions by automatically assembling new, or expanding existing\nmodels using published literature. Our approach integrates information\nextraction and clustering with simulation and formal analysis to allow for an\nautomated iterative process that includes assembling, testing and selecting the\nmost relevant models, given a set of desired system properties. We applied our\nmethodology to a model of the circuitry that con-trols T cell differentiation.\nTo evaluate our approach, we compare the model that we obtained, using our\nautomated model extension approach, with the previously published manually\nextended T cell differentiation model. Besides demonstrating automated and\nrapid reconstruction of a model that was previously built manually, ACCORDION\ncan assemble multiple models that satisfy desired properties. As such, it\nreplaces large number of tedious or even imprac-tical manual experiments and\nguides alternative hypotheses and interventions in biological systems.\n", " In recent years the coincidence of the operator relations equivalence after\nextension (EAE) and Schur coupling (SC) was settled for the Hilbert space case.\nFor Banach space operators, it is known that SC implies EAE, but the converse\nimplication is only known for special classes of operators, such as Fredholm\noperators with index zero and operators that can in norm be approximated by\ninvertible operators. In this paper we prove that the implication EAE\n$\\Rightarrow$ SC also holds for inessential Banach space operators. The\ninessential operators were introduced as a generalization of the compact\noperators, and include, besides the compact operators, also the strictly\nsingular and strictly co-singular operators; in fact they form the largest\nideal such that the invertible elements in the associated quotient algebra\ncoincide with (the equivalence classes of) the Fredholm operators.\n", " A novel directional coupler with a left-handed material (LHM) layer between\ntwo single-mode waveguides of usual material is introduced. The coupling system\nis analyzed with the supermode theory. It is shown that such a LHM layer of\nfinite length can shorten significantly the coupling length for the two\nsingle-mode waveguides. A LHM layer with two slowly tapered ends is used to\navoid the reflection loss at the ends.\n", " In this paper we use considerations of non-commutative geometry to deduce a\nmodel for QCD interactions. The model also explains within the same theoretical\nframework hitherto purely phenomenological characteristics of the quarks like\ntheir fractional charge, mass,handedness and confinement.\n", " The Euclidean-time hadronic current correlation functions, $G_P(\\tau, T)$ and\n$G_V(\\tau, T)$, of pseudoscalar and vector currents have recently been\ncalculated in lattice simulations of QCD and have been used to obtain the\ncorresponding spectral functions. We have used the Nambu-Jona-Lasinio (NJL)\nmodel to calculate such spectral functions, as well as the Euclidean-time\ncorrelators, and have made a comparison to the lattice results for the\ncorrelators. We find evidence for the type of temperature dependence of the NJL\ncoupling parameters that we have used in previous studies of the mesonic\nconfinement-deconfinement transition. We also see that the spectral functions\nobtained when using the maximum-entropy-method (MEM) and the lattice data\ndiffer from the spectral functions that we calculate in our chiral model.\nHowever, our results for the Euclidean-time correlators are in general\nagreement with the lattice results, with better agreement when our\ntemperature-dependent coupling parameters are used than when\ntemperature-independent parameters are used for the NJL model. We also discuss\nsome additional evidence for the utility of temperature-dependent coupling\nparameters for the NJL model. For example, if the constituent quark mass at T=0\nis $352 {MeV}$ in the chiral limit, the transition temperature is $T_c=208\n{MeV}$ for the NJL model with a standard momentum cutoff parameter. (If a\nGaussian momentum cutoff is used, we find $T_c=225 {MeV}$ in the chiral limit,\nwith $m=368 {MeV}$ at T=0.) The introduction of a weak temperature dependence\nfor the coupling constant will move the value of $T_c$ into the range 150-170\nMeV, which is more in accord with what is found in lattice simulations of QCD\nwith dynamical quarks.\n", " Tessellations are an important tool to model the microstructure of cellular\nand polycrystalline materials. Classical tessellation models include the\nVoronoi diagram and Laguerre tessellation whose cells are polyhedra. Due to the\nconvexity of their cells, those models may be too restrictive to describe data\nthat includes possibly anisotropic grains with curved boundaries. Several\ngeneralizations exist. The cells of the generalized balanced power diagram are\ninduced by elliptic distances leading to more diverse structures. So far,\nmethods for computing the generalized balanced power diagram are restricted to\ndiscretized versions in the form of label images. In this work, we derive an\nanalytic representation of the vertices and edges of the generalized balanced\npower diagram in 2d. Based on that, we propose a novel algorithm to compute the\nwhole diagram.\n", " A characterization of instanton contributions to gauge field theory is given.\nThe dynamics of instantons with the gluon field is given in terms of\n'classical' instanton contributions, i.e. on the same footing as tree\namplitudes in field theory. This paramaterization depends on the kinematics of\nthe local instanton operators and their coupling dependence. A primary\nadvantage of this 'classical' formulation is that the weak-strong duality and\nits relations with the perturbative sector are possibly made more manifest.\n", " Six families of generalized hypergeometric series in a variable $x$ and an\narbitrary number of parameters are considered. Each of them is indexed by an\ninteger $n$. Linear recurrence relations in $n$ relate these functions and\ntheir product by the variable $x$. We give explicit factorizations of these\nequations as products of first order recurrence operators. Related recurrences\nare also derived for the derivative with respect to $x$. These formulas\ngeneralize well-known properties of the classical orthogonal polynomials.\n" ]

[ " The concept of microstructure of the real space, considered as a mathematical\nlattice of cells (the tessel-lattice), and notions of canonical particles and\nfields, which are generated by the space, are analyzed. Submicroscopic\nmechanics based on this concept is discussed and employed for in-depth study of\nthe nucleon-nucleon interaction. It is argued that a deformation coat is\ndeveloped in the real space around the nucleon (as is the case with any other\ncanonical particle such as electron, muon, etc.) and that they are the\ndeformation coats that are responsible for the appearance of nuclear forces.\nOne more source of nuclear forces is associated with inerton clouds,\nexcitations of the space tessel-lattice (the excitations are a substructure of\nnucleons' matter waves), which accompany moving nucleons as in the case of any\nother canonical particles. Two nuclear systems are under consideration: the\ndeuteron and a weight nucleus. It is shown that a weight nucleus is a cluster\nof interacting protons and neutrons. The condition of the cluster stability is\nobtained in the framework of the statistical mechanical approach. On the whole,\nthe paper proposes a radically new approach to a very old unsolved problem, the\norigin of the nuclear forces, allowing the understanding mechanism(s) occurring\nat low energy nuclear transmutations.\n", " The latest research of the proportionality of atomic weights of chemical\nelements made it possible to obtain 3 x 3 matrices for the calculation of\ninformation coefficients of proportionality Ip that can be used for 3D modeling\nof the structure of atom nucleus. The results of computer simulation show high\npotential of nucleus structure research for the characterization of their\nchemical and physical properties.\n", " Understanding the nature of the matter comprising the Solar System is crucial\nfor understanding the mechanism that generates the Earth's geomagnetic field\nand the magnetic fields of other planets and satellites. The commonality in the\nSolar System of matter like that of the inside of the Earth, together with\ncommon nuclear reactor operating conditions,forms the basis for generalizing\nthe author's concept of nuclear geomagnetic field generation to planetary\nmagnetic field generation by natural planetocentric nuclear fission reactors.\n", " The surface quantum Hall state, magneto-electric phenomena and their\nconnection to axion electrodynamics have been studied intensively for\ntopological insulators. One of the obstacles for observing such effects comes\nfrom nonzero conductivity of the bulk. To overcome this obstacle we propose to\nuse an external magnetic field to suppress the conductivity of the bulk\ncarriers. The magnetic field dependence of galvanomagnetic and electromagnetic\nresponses of the whole system shows anomalies due to broken time-reversal\nsymmetry of the surface quantum Hall state, which can be used for its\ndetection. In particular, we find linear bulk dc magnetoresistivity and a\nquadratic field dependence of the Hall angle, shifted rf cyclotron resonance,\nnonanalytic microwave transmission coefficient and saturation of the Faraday\nrotation angle with increasing magnetic field or wave frequency.\n", " The purpose of this contribution is to show how a nuclear field theory\nfollows naturally from the structure of four-dimensional Riemannian geometry. A\nYang-Mills field is introduced by constructing fibres that include all possible\nexchanges of spin, parity and charge such that the collective quantum numbers\nremain the same. In this way O(4) internal symmetry transformations are found\nand a connection is obtained by exponentiation of a CP-invariant operator C\nassociated with the ground state. The metric is Calabi-Yau and Einstein. Carbon\n13 is chosen as an example because it is the lightest nucleus to exhibit small\nspin mutations even though there is no deformation parameter in the O(4)\ncommutation relations. Instead a supersymmetric transformation replaces a\nquantum group. Mirror symmetry is also discussed.\n", " A model-independent treatment of dark-matter particle elastic scattering has\nbeen developed, yielding the most general interaction for WIMP-nucleon\nlow-energy scattering, and the resulting amplitude has been embedded in the\nnucleus, taking into account the selection rules imposed by parity and\ntime-reversal. One finds that, in contrast to the usual\nspin-independent/spin-dependent (SI/SD) formulation, the resulting cross\nsection contains six independent nuclear response functions, three of which are\nassociated with possible velocity-dependent interactions. We find that current\nexperiments are four orders of magnitude more sensitive to derivative couplings\nthan is apparent in the standard SI/SD treatment, which necessarily associates\nsuch interactions with cross sections proportional to the square of the WIMP\nvelocity relative to the nuclear center of mass.\n", " We propose a new approach to explore CPT violation of neutrino oscillation\nthrough a fluctuating matter based on time-dependent geometric quantities. By\nmapping the neutrino oscillation onto a Poincare sphere structure, we obtain an\nanalytic solution of master equation and further define the geometric\nquantities, i.e., radius of Poincare sphere and geometric phase. We find that\nthe mixing process between electron and muon neutrinos can be described by the\nradius of Poincare sphere that depends on the intrinsic CP-violating angle.\nSuch a radius reveals a dynamic mechanism of CPT-violation, i.e., both\nspontaneous symmetry breaking and Majorana-Dirac neutrino confusion. We show\nthat the time-dependent geometric phase can be used to find the neutrino nature\nand observe the CPT-violation because it is strongly enhanced under the\nneutrino propagation. We further show that the time-dependent geometric phase\ncan be easily detected by simulating the neutrino oscillation based on\nfluctuating magnetic fields in nuclear magnetic resonance, which makes the\nexperimental observation of CPT-violation possible in the neutrino mixing and\noscillation.\n", " We characterize the measurement complexity of compressed sensing of signals\ndrawn from a known prior distribution, even when the support of the prior is\nthe entire space (rather than, say, sparse vectors). We show for Gaussian\nmeasurements and \\emph{any} prior distribution on the signal, that the\nposterior sampling estimator achieves near-optimal recovery guarantees.\nMoreover, this result is robust to model mismatch, as long as the distribution\nestimate (e.g., from an invertible generative model) is close to the true\ndistribution in Wasserstein distance. We implement the posterior sampling\nestimator for deep generative priors using Langevin dynamics, and empirically\nfind that it produces accurate estimates with more diversity than MAP.\n", " The rapid development in information technology has immensely contributed to\nthe use of modern approaches for visualizing volumetric data. Consequently,\nmedical volume visualization is increasingly attracting attention towards\nachieving an effective visualization algorithm for medical diagnosis and\npre-treatment planning. Previously, research has been addressing implementation\nof algorithm that can visualize 2-D images into 3-D. Meanwhile, achieving such\na rendering algorithm at an interactive speed and of good robustness to handle\nmass data still remains a challenging issue. However, in medical diagnosis,\nfinding the exact location of brain tumor or diseases is an important step of\nsurgery / disease management. This paper proposes a GPU-based raycasting\nalgorithm for accurate allocation and localization of human brain abnormalities\nusing magnetic resonance (MRI) images of normal and abnormal patients.\n Keywords: Brain tumor, graphic processing units, magnetic resonance imaging,\nvolume visualization\n", " The phase diagram of water has been calculated for the TIP4PQ/2005 model, an\nempirical rigid non-polarisable model. The path integral Monte Carlo technique\nwas used, permitting the incorporation of nuclear quantum effects. The\ncoexistence lines were traced out using the Gibbs-Duhem integration method,\nonce having calculated the free energies of the liquid and solid phases in the\nquantum limit, which were obtained via thermodynamic integration from the\nclassical value by scaling the mass of the water molecule. The resulting phase\ndiagram is qualitatively correct, being displaced to lower temperatures by\n15-20K. It is found that the influence of nuclear quantum effects are\ncorrelated to the tetrahedral order parameter.\n", " The {\\em compressed stack} is a data structure designed by Barba {\\em et al.}\n(Algorithmica 2015) that allows to reduce the amount of memory needed by an\nalgorithm (at the cost of increasing its runtime). In this paper we introduce\nthe first implementation of this data structure and make its source code\npublicly available.\n Together with the implementation we analyze the performance of the compressed\nstack. In our synthetic experiments, considering different test scenarios and\nusing data sizes ranging up to $2^{30}$ elements, we compare it with the\nclassic (uncompressed) stack, both in terms of runtime and memory used.\n Our experiments show that the compressed stack needs significantly less\nmemory than the usual stack (this difference is significant for inputs\ncontaining $2000$ or more elements). Overall, with a proper choice of\nparameters, we can save a significant amount of space (from two to four orders\nof magnitude) with a small increase in the runtime ($2.32$ times slower on\naverage than the classic stack). These results holds even in test scenarios\nspecifically designed to be challenging for the compressed stack.\n", " Different space-filling curves (SFCs) are briefly reviewed in this paper, and\na new one is proposed. A century has passed between the inception of this kind\nof curves, since then they have been found useful in computer science,\nparticularly in data storage and indexing due to their clustering properties,\nbeing Hilbert curve the most well-known member of the family of fractals. The\nproposed Aztec curve, with similar characteristics to the Hilbert's curve, is\nintroduced in this paper, accompanied by a grammatical description for its\nconstruction. It yields the possibility of creating bi-dimensional clusters,\nnot available for Hilbert nor Peano curves. Additional to this, a case of\napplication on the scope of Compressed Sensing is implemented, in which the use\nof Hilbert curve is contrasted with Aztec curve, having a similar performance,\nand positioning the Aztec curve as viable and a new alternative for future\nexploitation on applications that make use of SFC's.\n", " We exploit magnetostriction in polydisperse ferrofluids in order to generate\nnonlinear responses, and apply a thermodynamical method to derive the desired\nnonlinear magnetic susceptibility. For an ideal gas, this method has been\ndemonstrated to be in excellent agreement with a statistical method. In the\npresence of a sinusoidal ac magnetic field, the magnetization of the\npolydisperse ferrofluid contains higher-order harmonics, which can be extracted\nanalytically by using a perturbation approach. We find that the harmonics are\nsensitive to the particle distribution and the degree of field-induced\nanisotropy of the system. In addition, we find that the magnetization is higher\nin the polydisperse system than in the monodisperse one, as also found by a\nrecent Monte Carlo simulation. Thus, it seems possible to detect the size\ndistribution in a polydisperse ferrofluid by measuring the harmonics of the\nmagnetization under the influence of magnetostriction.\n", " In this paper we develop a general theory of compressed sensing for analog\nsignals, in close similarity to prior results for vectors in finite dimensional\nspaces that are sparse in a given orthonormal basis. The signals are modeled by\nfunctions in a reproducing kernel Hilbert space. Sparsity is defined as the\nminimal number of terms in expansions based on the kernel functions. Minimizing\nthis number is under certain conditions equivalent to minimizing an atomic\nnorm, the pre-dual of the supremum norm for functions in the Hilbert space. The\nnorm minimizer is shown to exist based on a compactness argument. Recovery\nbased on minimizing the atomic norm is robust and stable, so it provides\ncontrollable accuracy for recovery when the signal is only approximately sparse\nand the measurement is corrupted by noise.\n As applications of the theory, we include results on the recovery of sparse\nbandlimited functions and functions that have a sparse inverse short-time\nFourier transform.\n", " New macroscopic quantum interference effects in the behavior of the\nantiferromagnetic nanoclusters under action of a swept magnetic field are\npredicted and theoretically investigated, namely : oscillations of the magnetic\nsusceptibility of nanocluster similar to the electron Bloch oscillations in\ncrystal and the corresponding Stark ladder-like resonances in ac-magnetic\nfield. These effects are related to that of the macroscopic quantum coherence\nphenomena. The analogy of the antiferromagnetic nanoclusters quantum behavior\nand Josephson junction with a small capacity was revealed as well.\n", " We employ a propagator technique to derive a new relativistic $1/\\qq$\nexpansion of the structure function of a nucleus, composed of point-nucleons.\nWe exploit non-relativistic features of low-momentum nucleons in the target and\nonly treat relativistically the nucleon after absorption of a high-momentum\nvirtual photon. The new series permits a 3-dimensional reduction of each term\nand a formal summation of all Final State Interaction terms. We then show that\na relativistic structure function can be obtained from its non-relativistic\nanalog by a mere change of a scaling variable and an addition of an energy\nshift. We compare the obtained result with an ad hoc generalized\nGersch-Rodriguez-Smith theory, previously used in computations of nuclear\nstructure functions.\n", " Here, we investigate spin Hall angle of Pt in Ni80Fe20/Pt bilayer system by\nusing a broadband spin pumping and inverse spin Hall effect measurement. An\nout-of-plane excitation geometry with application of external magnetic field\nperpendicular to the charge current direction is utilized in order to suppress\nunwanted galvanomagnetic effects. Magnetization precession angle on\nferromagnetic resonance for wide excitation frequency range (4-14 GHz) is\nestimated from the rectification voltage of anisotropic magnetoresistance (AMR)\nand a conventional method of using microwave power in a coplanar waveguide. A\nmarked difference in the precession angle profiles for the different methods is\nobserved, resulting in the large variation in estimated values of spin current\ndensity at Ni80Fe20/Pt interface. The frequency dependence of the spin current\ndensity estimated using AMR effect is found to be similar to that of the\ninverse spin Hall voltage. We obtain the frequency-invariant spin Hall angle of\n0.067.\n", " Using numerical simulations of quenched SU(2) gauge theory we demonstrate\nthat an external magnetic field leads to spontaneous generation of quark\ncondensates with quantum numbers of electrically charged rho mesons if the\nstrength of the magnetic field exceeds the critical value eBc = 0.927(77) GeV^2\nor Bc =(1.56 \\pm 0.13) 10^{16} Tesla. The condensation of the charged rho\nmesons in strong magnetic field is a key feature of the magnetic-field-induced\nelectromagnetic superconductivity of the vacuum.\n", " Ab initio QED calculations of the nuclear magnetic shielding constant in\nhelium-like ions are presented. We combine the nonrelativistic QED approach\nbased on an expansion in powers of the fine-structure constant $\\alpha$ and the\nso-called ``all-order'' QED approach which includes all orders in the parameter\n$Z\\alpha$ but uses a perturbation expansion in the parameter $1/Z$ (where $Z$\nis the nuclear charge number). The combination of the two complementary methods\nmakes our treatment applicable both to low-$Z$ and high-$Z$ ions. Our\ncalculations confirm the presence of a rare antiscreening effect for the\nrelativistic shielding correction and demonstrate the importance of the\ninclusion of the negative-energy part of the Dirac spectrum.\n", " Radical pair recombination reactions are known to be sensitive to the\napplication of both low and high magnetic fields. The application of a weak\nmagnetic field reduces the singlet yield of a singlet-born radical pair,\nwhereas the application of a strong magnetic field increases the singlet yield.\nThe high field effect arises from energy conservation: when the magnetic field\nis stronger than the sum of the hyperfine fields in the two radicals, ${\\rm\nS}\\to {\\rm T}_{\\pm}$ transitions become energetically forbidden, thereby\nreducing the number of pathways for singlet to triplet interconversion. The low\nfield effect arises from symmetry breaking: the application of a weak magnetic\nfield lifts degeneracies among the zero field eigenstates and increases the\nnumber of pathways for singlet to triplet interconversion. However, the details\nof this effect are more subtle, and have not previously been properly\nexplained. Here we present a complete analysis of the low field effect in a\nradical pair containing a single proton, and in a radical pair in which one of\nthe radicals contains a large number of hyperfine-coupled nuclear spins. We\nfind that the new transitions that occur when the field is switched on are\nbetween ${\\rm S}$ and ${\\rm T}_0$ in both cases, and not between ${\\rm S}$ and\n${\\rm T}_{\\pm}$ as has previously been claimed. We then illustrate this result\nby using it in conjunction with semiclassical spin dynamics simulations to\naccount for the observation of a biphasic--triphasic--biphasic transition with\nincreasing magnetic field strength in the magnetic field effect on the\ntime-dependent survival probability of a photoexcited\ncarotenoid-porphyrin-fullerene radical pair.\n", " The experimental study of hydrogen bonds and their symmetrisation under\nextreme conditions is predominantly driven by diffraction methods, despite\nchallenges of localising or probing the hydrogen subsystems directly. Until\nrecently, H-bond symmetrisation has been addressed in terms of either nuclear\nquantum effects, spin crossovers or direct structural transitions; often\nleading to contradictory interpretations when combined. Here, we present\nhigh-resolution \\textit{in-situ} $^1$H-NMR experiments in diamond anvil cells\ninvestigating a wide range of hydrogen bonded systems at pressure ranges of up\nto 90 GPa covering their respective H-bond symmetrisation. We found pronounced\nminima in the pressure dependence of the NMR resonance line-widths associated\nwith a maximum in hydrogen mobility, precursor to a localisation of hydrogen\natoms. These minima, independent of of the chemical environment of the linear O\n-- H-O unit, can be found in a narrow range of oxygen-oxygen distances between\n2.44 and 2.45 \\AA, leading to an average critical oxygen-oxygen distance of\n$\\bar{r}_{\\rm OO}^{crit}=2.443(1)$ \\AA.\n", " Matrix data sets are common nowadays like in biomedical imaging where the\nDiffusion Tensor Magnetic Resonance Imaging (DT-MRI) modality produces data\nsets of\n 3D symmetric positive definite matrices anchored at voxel positions capturing\nthe anisotropic diffusion properties of water molecules in biological tissues.\nThe space of symmetric matrices can be partially ordered using the L\\\"owner\nordering, and computing extremal matrices dominating a given set of matrices is\na basic primitive used in matrix-valued signal processing. In this letter, we\ndesign a fast and easy-to-implement iterative algorithm to approximate\narbitrarily finely these extremal matrices. Finally, we discuss on extensions\nto matrix clustering.\n", " Formed in the aftermath of gravitational core-collapse supernova explosions,\nneutron stars are unique cosmic laboratories for probing the properties of\nmatter under extreme conditions that cannot be reproduced in terrestrial\nlaboratories. The interior of a neutron star, endowed with the highest magnetic\nfields known and with densities spanning about ten orders of magnitude from the\nsurface to the centre, is predicted to exhibit various phases of dense strongly\ninteracting matter, whose physics is reviewed in this chapter. The outer layers\nof a neutron star consist of a solid nuclear crust, permeated by a neutron\nocean in its densest region, possibly on top of a nuclear \"pasta\" mantle. The\nproperties of these layers and of the homogeneous isospin asymmetric nuclear\nmatter beneath constituting the outer core may still be constrained by\nterrestrial experiments. The inner core of highly degenerate, strongly\ninteracting matter poses a few puzzles and questions which are reviewed here\ntogether with perspectives for their resolution. Consequences of the\ndense-matter phases for observables such the neutron-star mass-radius\nrelationship and the prospects to uncover their structure with modern\nobservational programmes are touched upon.\n", " Visible light, ultraviolet and x-ray radiation have been found to increase\nchemical reactivity of water. The irradiated solution of water in acetonitrile\nreacts with triethyl phosphite considerably faster than the non-irradiated\ncontrol solution. This phenomenon is accounted for by the decomposition of\nwater clusters under the influence of light with the formation of chemically\nmore active free water molecules.\n", " Standard lore holds that magnetic forces are incapable of doing mechanical\nwork. More precisely, the claim is that whenever it appears that a magnetic\nforce is doing work, the work is actually being done by another force, with the\nmagnetic force serving only as an indirect mediator. However, the most familiar\ninstances of magnetic forces acting in everyday life, such as when bar magnets\nlift other bar magnets, appear to present manifest evidence of magnetic forces\ndoing work. These sorts of counterexamples are often dismissed as arising from\nquantum effects that lie outside the classical regime. In this paper, we show\nthat quantum theory is not needed to account for these phenomena, and that\nclassical electromagnetism admits a model of elementary magnetic dipoles on\nwhich magnetic forces can indeed do work. In order to develop this model, we\nrevisit the foundational principles of the classical theory of\nelectromagnetism, showcase the importance of constraints from relativity,\nexamine the structure of the multipole expansion, and study the connection\nbetween the Lorentz force law and conservation of energy and momentum.\n", " Compton scattering is a key process shaping spectra formation and accretion\nflow dynamics in accreting strongly magnetized neutron stars. A strong magnetic\nfield affects the scattering cross section and makes it dependent on photon\nenergy, momentum, and polarization state. Using Monte Carlo simulations, we\ninvestigate statistical features of Compton scattering of polarized X-ray\nradiation in a strong magnetic field. Our analysis is focused on photon gas\nbehaviour well inside the scattering region. We take into account the resonant\nscattering at the fundamental cyclotron frequency, thermal distribution of\nelectrons at the ground Landau level, and bulk velocity of the electron gas. We\nshow that (i) the photons scattered around the cyclotron energy by the electron\ngas at rest tend to acquire the final energy close to the cyclotron one with a\nvery small dispersion measure; (ii) the redistribution of photons within the\nDoppler core of cyclotron resonance differs significantly from the complete\nredistribution; (iii) the efficiency of momentum transfer from photons to the\nelectron gas is affected by the temperature of electron gas both for photons at\ncyclotron energy and below it; (iv) the momentum transfer from photons to the\nelectron gas of non-zero bulk velocity is more efficient in the case of\nmagnetic scattering.\n", " The problems considered refer to the material equations of electric- and\nmagnetoelectric induction. Some contradictions found in fundamental studies on\nclassical electrodynamics have been explained. The notion magnetoelectric\ninduction has been introduced, which permits symmetrical writing of the\ninduction laws. It is shown that the results of the special theory of\nrelativity can be obtained from these laws through the Galilean\ntransformations. The permittivity and permeability of materials media are shown\nto be independent of frequency. The notions magnetoelectrokinetic and\nelectromagnetopotential waves and kinetic capacity have been introduced. It is\nshown that along with the longitudinal Langmuir resonance, the transverse\nresonance is possible in nonmagnetized plasma, and both the resonances are\ndegenerate. A new notion scalar-vector potential is introduced, which permits\nsolution of all present-day problems of classical electrodynamics. The use of\nthe scalar-vector potential makes the magnetic field notion unnecessary.\n", " We present the first sublinear memory sketch that can be queried to find the\nnearest neighbors in a dataset. Our online sketching algorithm compresses an N\nelement dataset to a sketch of size $O(N^b \\log^3 N)$ in $O(N^{(b+1)} \\log^3\nN)$ time, where $b < 1$. This sketch can correctly report the nearest neighbors\nof any query that satisfies a stability condition parameterized by $b$. We\nachieve sublinear memory performance on stable queries by combining recent\nadvances in locality sensitive hash (LSH)-based estimators, online kernel\ndensity estimation, and compressed sensing. Our theoretical results shed new\nlight on the memory-accuracy tradeoff for nearest neighbor search, and our\nsketch, which consists entirely of short integer arrays, has a variety of\nattractive features in practice. We evaluate the memory-recall tradeoff of our\nmethod on a friend recommendation task in the Google Plus social media network.\nWe obtain orders of magnitude better compression than the random projection\nbased alternative while retaining the ability to report the nearest neighbors\nof practical queries.\n", " We elucidate the nature of neutral collective excitations of fractional\nquantum Hall liquids in the long-wavelength limit. We demonstrate that they are\nchiral gravitons carrying angular momentum - 2, which are quanta of quantum\nmotion of an internal metric, and show up as resonance peaks in the systems\nresponse to what is the fractional Hall analog of gravitational waves. Relation\nwith existing and possible future experimental work that can detect these\nfractional quantum Hall gravitons and reveal their chirality are discussed.\n", " Ab initio molecular dynamics (AIMD) is a powerful tool to predict properties\nof molecular and condensed matter systems. The quality of this procedure is\nbased on accurate electronic structure calculations. The development of quantum\nprocessors has shown great potential for the efficient evaluation of accurate\nground and excited state energies of molecular systems, opening up new avenues\nfor molecular dynamics simulations. In this work we address the use of\nvariational quantum algorithms for the calculation of accurate atomic forces to\nbe used in AIMD. In particular, we provide solutions for the alleviation of the\nstatistical noise associated to the measurements of the expectation values of\nenergies and forces, as well as schemes for the mitigation of the hardware\nnoise sources (in particular, gate infidelities, qubit decoherence and readout\nerrors). Despite the relative large error in the calculation of the potential\nenergy, our results show that the proposed algorithms can provide reliable MD\ntrajectories in the microcanonical (constant energy) ensemble. Further,\nexploiting the intrinsic noise arising from the quantum measurement process, we\nalso propose a Langevin dynamics algorithm for the simulation of canonical,\ni.e., constant temperature, dynamics. Both algorithms (microcanonical and\ncanonical) are applied to the simulation of simple molecular systems such as H2\nand H3+. Finally, we also provide results for the dynamics of H2 obtained with\nIBM quantum computer ibmq_athens.\n", " The periodically kicked quantum rotor is known for non-classical effects such\nas quantum localisation in angular momentum space or quantum resonances in\nrotational excitation. These phenomena have been studied in diverse systems\nmimicking the kicked rotor, such as cold atoms in optical lattices, or coupled\nphotonic structures. Recently, it was predicted that several solid state\nquantum localisation phenomena - Anderson localisation, Bloch oscillations, and\nTamm-Shockley surface states - may manifest themselves in the rotational\ndynamics of laser-kicked molecules. Here, we report the first observation of\nrotational Bloch oscillations in a gas of nitrogen molecules kicked by a\nperiodic train of femtosecond laser pulses. A controllable detuning from the\nquantum resonance creates an effective accelerating potential in angular\nmomentum space, inducing Bloch-like oscillations of the rotational excitation.\nThese oscillations are measured via the temporal modulation of the refractive\nindex of the gas. Our results introduce room-temperature laser-kicked molecules\nas a new laboratory for studies of localisation phenomena in quantum transport.\n", " It is shown that the strong enhancement of the Earth (day-night) effect for\nsolar neutrinos crossing the Earth core in the case of the small mixing angle\nMSW electron neutrino to muon (tau) neutrino transition solution of the solar\nneutrino problem is due to a new resonance effect in the solar neutrino\ntransitions in the Earth and not just to the MSW effect in the core. The effect\nis in many respects similar to the electron paramagnetic resonance. The\nconditions for existence of this new resonance effect are discussed. They\ninclude specific constraints on the neutrino oscillation lengths in the Earth\nmantle and in the Earth core, thus the resonance is a ``neutrino oscillation\nlength resonance''. The effect exhibits strong dependence on the neutrino\nenergy. Analytic expression for the probability accounting for the solar\nneutrino transitions in the Earth, which provides a high precision description\nof the transitions, including the new resonance effect, is derived. The\nimplications of our results for the searches of the day-night asymmetry in the\nsolar neutrino experiments are briefly discussed. The new resonance effect is\noperative also in the muon neutrino to electron neutrino (electron neutrino to\nmuon neutrino) transitions of atmospheric neutrinos crossing the Earth core.\n", " The effect of signal integration through cis-regulatory modules (CRMs) on\nsynchronization and clustering of populations of two-component genetic\noscillators coupled by quorum sensing is in detail investigated. We find that\nthe CRMs play an important role in achieving synchronization and clustering.\nFor this, we investigate 6 possible cis-regulatory input functions (CRIFs) with\nAND, OR, ANDN, ORN, XOR, and EQU types of responses in two possible kinds of\ncell-to-cell communications: activator-regulated communication (i.e., the\nautoinducer regulates the activator) and repressor-regulated communication\n(i.e., the autoinducer regulates the repressor). Both theoretical analysis and\nnumerical simulation show that different CRMs drive fundamentally different\ncellular patterns, such as complete synchronization, various cluster-balanced\nstates and several cluster-nonbalanced states.\n", " The nuclear matrix element determining the $pp\\to d e^+ \\nu$ fusion cross\nsection and the Gamow-Teller matrix element contributing to tritium\n$\\beta$-decay are calculated with lattice Quantum Chromodynamics (QCD) for the\nfirst time. Using a new implementation of the background field method, these\nquantities are calculated at the SU(3)-flavor-symmetric value of the quark\nmasses, corresponding to a pion mass of $m_\\pi$ ~ 806 MeV. The Gamow-Teller\nmatrix element in tritium is found to be 0.979(03)(10) at these quark masses,\nwhich is within $2\\sigma$ of the experimental value. Assuming that the\nshort-distance correlated two-nucleon contributions to the matrix element\n(meson-exchange currents) depend only mildly on the quark masses, as seen for\nthe analogous magnetic interactions, the calculated $pp\\to d e^+ \\nu$\ntransition matrix element leads to a fusion cross section at the physical quark\nmasses that is consistent with its currently accepted value. Moreover, the\nleading two-nucleon axial counterterm of pionless effective field theory is\ndetermined to be $L_{1,A}=3.9(0.1)(1.0)(0.3)(0.9)\\ {\\rm fm}^3$ at a\nrenormalization scale set by the physical pion mass, also in agreement with the\naccepted phenomenological range. This work concretely demonstrates that weak\ntransition amplitudes in few-nucleon systems can be studied directly from the\nfundamental quark and gluon degrees of freedom and opens the way for subsequent\ninvestigations of many important quantities in nuclear physics.\n", " Inherent asymmetry in the tunneling barriers of few-electron quantum dots\ninduces intrinsically different tunneling currents for forward and reverse\nsource-drain biases in the non-linear transport regime. Here we show that in\naddition to spin selection rules, overlap matrix elements between many-body\nstates are crucial for the correct description of tunneling transmission\nthrough quantum dots at large magnetic fields. Signatures of excited\n(N-1)-electron states in the transport process through the N-electron system\nare clearly identified in the measured transconductances. Our analysis clearly\nconfirms the validity of single-electron quantum transport theory in quantum\ndots.\n", " A tuning of the light transmission properties of 1D photonic structures\nemploying an external stimulus is very attracting and opens the way to the\nfabrication of optical switches for colour manipulation in sensing, lighting,\nand display technology. We present the electric field-induced tuning of the\nlight transmission in a photonic crystal device, made by alternating layers of\nsilver nanoparticles and titanium dioxide nanoparticles. We show a shift of\naround 10 nm for an applied voltage of 10 V. We ascribe the shift to an\naccumulation of charges at the silver/TiO2 interface due to electric field,\nresulting in an increase of the number of charges contributing to the plasma\nfrequency in silver, giving rise to a blue shift of the silver plasmon band,\nwith concomitant blue shift of the photonic band gap. The employment of a\nrelatively low applied voltage gives the possibility to build a compact and\nlow-cost device. We also propose the fabrication of 1D photonic crystal and\nmicrocavities employing a magneto-optical material as TGG (Tb3Ga5O12). With\nthese structures we can observe a shift of 22 nm with a magnetic field of 5 T,\nat low temperature (8 K). The option to tune the colour of a photonic crystal\nwith magnetic field is interesting because of the possibility to realize\ncontactless optical switches. We also discuss the possibility to achieve the\ntuning of the photonic band gap with UV light in photonic crystals made with\nindium tin oxide (ITO).\n", " A generic and intuitive model for coherent energy transport in multiple\nminima systems coupled to a quantum mechanical bath is shown. Using a simple\nspin-boson system, we illustrate how a generic donor-acceptor system can be\nbrought into resonance using a narrow band of vibrational modes, such that the\ntransfer efficiency of an electron-hole pair (exciton) is made arbitrarily\nhigh. Coherent transport phenomena in nature are of renewed interest since the\ndiscovery that a photon captured by the light-harvesting complex (LHC) in\nphotosynthetic organisms can be conveyed to a chemical reaction centre with\nnear-perfect efficiency. Classical explanations of the transfer use stochastic\ndiffusion to model the hopping motion of a photo-excited exciton. This accounts\ninadequately for the speed and efficiency of the energy transfer measured in a\nseries of recent landmark experiments. Taking a quantum mechanical perspective\ncan help capture the salient features of the efficient part of that transfer.\nTo show the versatility of the model, we extend it to a multiple minima system\ncomprising seven-sites, reminiscent of the widely studied Fenna-Matthews-Olson\n(FMO) light-harvesting complex. We show that an idealised transport model for\nmultiple minima coupled to a narrow-band phonon can transport energy with\narbitrarily high efficiency.\n", " Electron transport in realistic physical and chemical systems often involves\nthe non-trivial exchange of energy with a large environment, requiring the\ndefinition and treatment of open quantum systems. Because the time evolution of\nan open quantum system employs a non-unitary operator, the simulation of open\nquantum systems presents a challenge for universal quantum computers\nconstructed from only unitary operators or gates. Here we present a general\nalgorithm for implementing the action of any non-unitary operator on an\narbitrary state on a quantum device. We show that any quantum operator can be\nexactly decomposed as a linear combination of at most four unitary operators.\nWe demonstrate this method on a two-level system in both zero and finite\ntemperature amplitude damping channels. The results are in agreement with\nclassical calculations, showing promise in simulating non-unitary operations on\nintermediate-term and future quantum devices.\n", " The effects of the first nonlinear corrections to the DGLAP equations are\nstudied in light of the HERA data. Saturation limits are determined in the\nDGLAP+GLRMQ approach for the free proton and for the Pb nucleus.\n", " Trapped atoms and ions are among the best controlled quantum systems which\nfind widespread applications in quantum information, sensing and metrology. For\nmolecules, however, a similar degree of control is currently lacking owing to\ntheir complex energy-level structure. Quantum-logic protocols in which atomic\nions serve as probes for molecular ions are a promising route for achieving\nthis level of control, especially with homonuclear molecules that decouple from\nblack-body radiation. Here, a quantum-non-demolition protocol on single trapped\nN$_2^+$ molecules is demonstrated. The spin-rovibronic state of the molecule is\ndetected with more than 99% fidelity and the position and strength of a\nspectroscopic transition in the molecule are determined, both without\ndestroying the molecular quantum state. The present method lays the foundations\nfor new approaches to molecular precision spectroscopy, for state-to-state\nchemistry on the single-molecule level and for the implementation of molecular\nqubits.\n", " Starting from the meaning of the conjugate of a complex Hilbert space,\nincluding a related application of the theorem of Fr\\'{e}chet-Riesz (by which\nan analysis of semilinear operators can be reduced to - linear - operator\ntheory) to a revisit of applications of nuclear and absolutely $p$-summing\noperators in algebraic quantum field theory in the sense of Araki, Haag and\nKastler ($p=2$) and more recently in the framework of general probabilistic\nspaces ($p=1$), we will outline that Banach operator ideals in the sense of\nPietsch, or equivalently tensor products of Banach spaces in the sense of\nGrothendieck are even lurking in the foundations and philosophy of quantum\nphysics and quantum information theory. In particular, we concentrate on their\nimportance in algebraic quantum field theory. In doing so, we establish a\ncanonical isometric isomorphism between the Hilbert spaces $H\\otimes_2 (K\n\\otimes_2 L)$ and $(H \\otimes_2 K) \\otimes_2 L$ (Theorem 3.8) and revisit the\nrole of trace class operators.\n A few applications are specified, including the appropriateness of the class\nof Hilbert-Schmidt operators and an implied Banach operator ideal\nrepresentation of the tensor product of two complex Hilbert spaces $H \\otimes_2\nK$ (Proposition 3.4) and a purely linear algebraic description of the quantum\nteleportation process (Example 3.10).\n", " Photothermal-Induced Resonance (PTIR) is increasingly used in the measurement\nof infrared absorption spectra of sub-micrometer objects. The technique\nmeasures IR absorption spectra by relying on the photothermal effect induced by\na rapid pulse of light and the excitation of the resonance spectrum of an AFM\ncantilever in contact with the sample. In this work we assess the spatial\nresolution and depth response of PTIR in resonant mode while systematically\nvarying the pulsing frequency of the excitation laser and the cantilever\nresonance. The spatial resolution shows a shallow linear dependence on the\ninverse of the pulse frequency, which rules out tip size as a limiting factor\nfor resolution in the frequency range under investigation. Measured resolution\nvalues are also one order of magnitude lower than in the thermal diffusion\nlimit, excluding thermal wave propagation as a limiting factor. We also show\nthat the pulsing frequency of the laser and choice of cantilever resonance\naffect the intensity of the signal and the surface selectivity in PTIR images,\nwith higher frequencies providing increased surface selectivity. The results\nconfirm a difference in signal generation between resonant PTIR and other\nphotothermal techniques and indicate that photothermal induced heating and\nexpansion cannot fully account for observed intensity and resolution.\n", " The most common sellar lesion is the pituitary adenoma, and sellar tumors are\napproximately 10-15% of all intracranial neoplasms. Manual slice-by-slice\nsegmentation takes quite some time that can be reduced by using the appropriate\nalgorithms. In this contribution, we present a segmentation method for\npituitary adenoma. The method is based on an algorithm that we have applied\nrecently to segmenting glioblastoma multiforme. A modification of this scheme\nis used for adenoma segmentation that is much harder to perform, due to lack of\ncontrast-enhanced boundaries. In our experimental evaluation, neurosurgeons\nperformed manual slice-by-slice segmentation of ten magnetic resonance imaging\n(MRI) cases. The segmentations were compared to the segmentation results of the\nproposed method using the Dice Similarity Coefficient (DSC). The average DSC\nfor all datasets was 75.92% +/- 7.24%. A manual segmentation took about four\nminutes and our algorithm required about one second.\n", " Next-nearest-neighbor Heisenberg chain plays important roles in solid state\nphysics, such as predicting exotic electric properties of two-dimensional\nmaterials or magnetic properties of organic compounds. Direct experimental\nstudies of the many-body electron systems or spin systems associating to these\nmaterials are challenging tasks, while optical simulation provides an effective\nand economical way for immediate observation. Comparing with bulk optics,\nintegrated optics are more of fascinating for steady, large scale and long-time\nevolution simulations. Photonic crystal is an artificial microstructure\nmaterial with multiple methods to tune the propagation properties, which are\nessential for various simulation tasks. Here we report for the first time an\nexperimental simulation of next-nearest-neighbor Heisenberg chain with an\nintegrated optical chip of photonic crystal waveguide array. The use of\nphotonic crystal enhances evanescent field thus allows coupling between\nnext-nearest-neighbor waveguides in such a planar waveguide array, without\nbreaking the weak coupling condition of the coupled mode equation.\nParticularly, similarities between the model and coherent light propagation\ncould reach 0.99 in numerical simulations and 0.89 in experiment. Localization\neffect induced by second-order coupling and coupling strengthening with\nincreasing wavelengths were also revealed in both numerical simulations and\nexperiments. The platform proposed here is compatible with mature complementary\nmetal oxide semiconductor technology thus possesses the potential for\nlarger-scale problems and photonic crystals further allows simulations of\nspecific target Hamiltonians.\n", " Mechanical displacements of a nanoelectromechanical system (NEMS) shift the\nelectron trajectories and hence perturb phase coherent charge transport through\nthe device. We show theoretically that in the presence of a magnetic feld such\nquantum-coherent displacements may give rise to an Aharonov-Bohm-type of\neffect. In particular, we demonstrate that quantum vibrations of a suspended\ncarbon nanotube result in a positive nanotube magnetoresistance, which\ndecreases slowly with the increase of temperature. This effect may enable one\nto detect quantum displacement fluctuations of a nanomechanical device.\n", " We investigate the possibility of a strongly correlated Fractional Quantum\nHall (FQH) state in bulk three dimensional isotropic (not layered) materials.\nWe find that a FQH state can exist at low densities only if it is accompanied\nby a staging transition in which the electrons re-organize themselves in\nlayers, perpendicular to the magnetic field, at distances of order the magnetic\nlength apart. The Hartree energy associated to the staging transition is\noff-set by the correlation Fock energy of the 3D FQH state. We obtain the phase\ndiagram of bulk electrons in a magnetic field subject to Coulomb interactions\nas a function of carrier density and lattice constant. At very low densities,\nthe 3D FQH state exhibits a transition to a 3D Wigner crystal state stabilized\nby phonon correlations.\n", " In this paper we study the identification of an event region $X$ within a\nlarger region $Y$, in which the sensors are distributed by a Poisson process of\ndensity $\\lambda$ to detect this event region, i.e., its boundary. The model of\nsensor is a 0-1 sensor that decides whether it lies in $X$ or not, and which\nmight be incorrect with probability $p$. It also collects information on the\n0-1 values of the neighbors within some distance $r$ and revises its decision\nby the majority vote of these neighbors. In the most general setting, we\nanalyze this simple majority vote scheme and derive some upper and lower bounds\non the expected number of misclassified sensors. These bounds depend on several\nsensing parameters of $p$, $r$, and some geometric parameters of the event\nregion $X$. By making some assumptions on the shape of $X$, we prove a\nsignificantly improved upper bound on the expected number of misclassified\nsensors; especially for convex regions with sufficiently round boundary, and we\nfind that the majority vote scheme performs well in the simulation rather than\nits theoretical upper bound.\n", " We study the equilibrium instability of chiral quarkonia in a plasma in the\npresence of constant magnetic and electric field and at finite axial chemical\npotential using AdS/CFT duality. The model in use is a supersymmetric QCD at\nlarge 't$\\,$Hooft coupling and number of colors. We show that the presence of\nthe magnetic field and the axial chemical potential even in the absence of the\nelectric field make the system unstable. In a gapped system, a stable/unstable\nequilibrium state phase transition is observed and the initial transition\namplitude of the equilibrium state to the non-equilibrium state is\ninvestigated. We demonstrate that at zero temperature and large magnetic field\nthe instability grows linearly by increasing the quarkonium binding energy. In\nthe constant electric and magnetic field, the system is in a equilibrium state\nif the Ohm's law and the chiral magnetic effect cancel their effects. This\nhappens in a sub-space of $(E,B,T,\\mu_5)$ space with constraint equation\n$\\sigma_B B =- \\sigma E$, where $\\sigma$ and $\\sigma_B$ are called electric and\nchiral magnetic conductivity, respectively. We analyze the decay rate of a\ngapless system when this constraint is slightly violated.\n", " The rigorous quantum mechanical description of the collective interaction of\nmany molecules with the radiation field is usually considered numerically\nintractable, and approximation schemes must be employed. Standard spectroscopy\nusually contains some levels of perturbation theory, but under strong coupling\nconditions, other approximations are used. A common approximation is the\n1-exciton model in which processes involving weak excitations are described\nusing a basis comprising of the ground state and singly excited states of the\nmolecule cavity-mode system. In another frequently used approximation in\nnumerical investigations, the electromagnetic field is described classically,\nand the quantum molecular subsystem is treated in the mean-field Hartree\napproximation with its wavefunction assumed to be a product of single\nmolecules' wavefunctions. The former disregards states that take long time to\npopulate, and is therefore essentially a short time approximation. The latter\nis not limited in this way, but by its nature, disregards some intermolecular\nand molecule-field correlations. In this work, we directly compare results\nobtained from these approximations when applied to several prototype problems\ninvolving the optical response of molecule-in-optical cavities systems. In\nparticular, we show that our recent model investigation (J. Chem. Phys. 157,\n114108 (2022)) of the interplay between the electronic strong coupling and\nmolecular nuclear dynamics using the truncated 1-exciton approximation agrees\nvery well with the semiclassical mean-field calculation.\n", " Capacitance-Voltage (C-V) & Ballistic Current- Voltage (I-V) characteristics\nof Double Gate (DG) Silicon-on- Insulator (SOI) Flexible FETs having sub 35nm\ndimensions are obtained by self-consistent method using coupled Schrodinger-\nPoisson solver taking into account the quantum mechanical effects. Although,\nATLAS simulations to determine current and other short channel effects in this\ndevice have been demonstrated in recent literature, C-V & Ballistic I-V\ncharacterizations by using self-consistent method are yet to be reported. C-V\ncharacteristic of this device is investigated here with the variation of bottom\ngate voltage. The depletion to accumulation transition point (i.e. Threshold\nvoltage) of the C-V curve should shift in the positive direction when the\nbottom gate is negatively biased and our simulation results validate this\nphenomenon. Ballistic performance of this device has also been studied with the\nvariation of top gate voltage.\n" ]

[ " Extra dimensions are introduced: 3 in Classical Mechanics and 6 in\nRelativistic Mechanics, which represent orientations, resulting from rotations,\nof a particle, described by quaternions, and leading to a 7-dimensional,\nrespectively 10-dimensional, world. The concept of spin is thus defined in both\ncases, which results to be different from angular momentum. It leads to the\nknown concept of Quantum Mechanical spin in both cases. In particular,\napplication of these new ideas to (Relativistic) Quantum Mechanics, leads to\nquantization of energy levels in free electrons, which could be in principle\ndetectable via their spectrum, confirming the theory.\n", " A new wave-particle non-dualistic interpretation for the quantum formalism is\npresented by proving that the Schr\\\"odinger wave function is an `{\\it\ninstantaneous resonant spatial mode}' in which the quantum particle moves. The\nprobabilities in quantum mechanics arise only for the observer's perspective\ndue to the nature of doing experiments but they do not exist in Nature. In\nother words, quantum mechanics itself is not a probabilistic theory. This\nview-point is proved by deriving the Born's rule by showing it to be equal to\nthe experimentally observed relative frequencies. Also, it is shown that the\nclassical and quantum mechanical times are one and the same. By identifying the\nalready inherently existing mechanism for the collapse of wave function within\nthe quantum formalism, the present interpretation naturally proves why the\nCopenhagen interpretation is very successful in explaining the experimental\noutcomes. The Young's double-slit experiment, the Wheeler's delayed choice\nexperiment and the Afshar's experiment are unambiguously explained at a single\nquantum level. Both `measurement problem' and `retrocausality' are absent in\nthe present non-dualistic interpretation of quantum mechanics.\n", " The geometry of the symplectic structures and Fubini-Study metric is\ndiscussed. Discussion in the paper addresses geometry of Quantum Mechanics in\nthe classical phase space. Also, geometry of Quantum Mechanics in the\nprojective Hilbert space has been discussed for the chosen Quantum states.\nSince the theory of classical gravity is basically geometric in nature and\nQuantum Mechanics is in no way devoid of geometry, the explorations pertaining\nto more and more geometry in Quantum Mechanics could prove to be valuable for\nlarger objectives such as understanding of gravity.\n", " The notion of nonlocality implicitly implies there might be some kind of\nspooky action at a distance in nature, however, the validity of quantum\nmechanics has been well tested up to now. In this work it is argued that the\nnotion of nonlocality is physically improper, the basic principle of locality\nin nature is well respected by quantum mechanics, namely, the uncertainty\nprinciple. We show that the quantum bound on the Clauser, Horne, Shimony, and\nHolt (CHSH) inequality can be recovered from the uncertainty relation in a\nmultipartite setting. We further argue that the super-quantum correlation\ndemonstrated by the nonlocal box is not physically comparable with the quantum\none. The origin of the quantum structure of nature still remains to be\nexplained, some post-quantum theory which is more complete in some sense than\nquantum mechanics is possible and might not necessarily be a hidden variable\ntheory.\n", " Quantum field theory currently has a single standard mathematical\ncharacterization (the Standard Model), but no single accepted conceptual\nframework to interpret the mathematics. Many of these conceptualizations rely\non intuitive concepts carried over from classical physics (such as \"particle\"\nand \"causality\"). In this paper, instead of relying on classical concepts we\nattempt to infer a conceptualization directly from the mathematics. This\nreconceptualization leads to a physical reinterpretation of the paths involved\nin the Standard Model's action integral as traces of an accumulative process\nthat occurs within an \"extraverse\", of which our observable spacetime is a\nsingle \"slice\". We briefly outline a rigorous mathematical model which\nvalidates the physical reinterpretation, and leads to predictions that are\npotentially verifiable by experiment. We contrast our model with other popular\ninterpretations of quantum mechanics. We describe some of the metaphysical\nconsequences of our proposed perspective. Finally we present some speculations\nabout an alternative approach that could possibly lead to a more satisfying,\nintuitively-graspable theory.\n", " We review our approach to quantum mechanics adding also some new interesting\nresults. We start by giving proof of two important theorems on the existence of\nthe and Clifford algebras. This last algebra gives proof of the von Neumann\nbasic postulates on the quantum measurement explaining thus in an algebraic\nmanner the wave function collapse postulated in standard quantum theory. In\nthis manner we reach the objective to expose a self-consistent version of\nquantum mechanics. We give proof of the quantum like Heisenberg uncertainty\nrelations, the phenomenon of quantum Mach Zender interference as well as\nquantum collapse in some cases of physical interest We also discuss the problem\nof time evolution of quantum systems as well as the changes in space location.\nWe also give demonstration of the Kocken-Specher theorem, and also we give an\nalgebraic formulation and explanation of the EPR . By using the same approach\nwe also derive Bell inequalities. Our formulation is strongly based on the use\nof idempotents that are contained in Clifford algebra. Their counterpart in\nquantum mechanics is represented by the projection operators that are\ninterpreted as logical statements, following the basic von Neumann results.\nUsing the Clifford algebra we are able to invert such result. According to the\nresults previously obtained by Orlov in 1994, we are able to give proof that\nquantum mechanics derives from logic. We show that indeterminism and quantum\ninterference have their origin in the logic.\n", " The apparent impossibility of extending non-relativistic quantum mechanics to\na relativistic quantum theory is shown to be due to the insufficient structural\nrichness of the field of complex numbers over which quantum mechanics is built.\nA new number system with the properties needed to support an inherently\nrelativistic quantum theory is brought to light and investigated to a point\nsufficient for applications.\n", " While quantum mechanics allows spooky action at a distance at the level of\nthe wave-function, it also respects locality since there is no instantaneous\npropagation of real physical effects. We show that this feature can be proved\nin the standard interpretation of quantum mechanics by a simple general result\ninvolving commuting Hermitian operators corresponding to distant (causally\ndisconnected) observables. This is reminiscent of satisfying the locality\ncondition in relativistic quantum field theories.\n", " A description of the motion in noninertial reference frames by means of the\ninclusion of high time derivatives is studied. Incompleteness of the\ndescription of physical reality is a problem of any theory, both in quantum\nmechanics and classical physics. The ``stability principle'' is put forward.\nWe~also provide macroscopic examples of noninertial mechanics and verify the\nuse of high-order derivatives as nonlocal hidden variables on the basis of the\nequivalence principle when acceleration is equal to the gravitational field.\nAcceleration in this case is a function of high derivatives with respect to\ntime. The~definition of dark metrics for matter and energy is presented to\nreplace the standard notions of dark matter and dark energy. In the Conclusion\nsection, problem symmetry is noted for noninertial mechanics.\n", " Discrepancies and accords between quantum (QM) and classical mechanics (CM)\nrelated to expectation values and periods are found for both the simple\nharmonic oscillator (SHO) and a free particle in a box (FPB), which may apply\ngenerally. These indicate non-locality is expected throughout QM. The FPB\nenergy states violate the Correspondence Principle. Previously unexpected\naccords are found and proven that the classical and quantum expectation values\nare the same for the expectation value of the second moment and the beat period\n(i.e. beats between the phases for adjoining energy states) for the SHO for all\nquantum numbers, n. However, for the FPB the beat periods differ significantly\nat small n. It is shown that a particle's velocity in an infinite square well\nvaries, no matter how wide the box, nor how far the particle is from the walls.\nThe quantum free particle variances share an indirect commonality with the\nAharonov-Bohm and Aharonov-Casher effects in that there is a quantum action in\nthe absence of a force. The concept of an \"Expectation Value over a Partial\nWell Width\" is introduced. This paper raises the question as to whether these\ninconsistencies are undetectable, or can be empirically ascertained. These\ninherent variances may need to be fixed, or nature is manifestly more\nnon-classical than expected.\n", " A brief history is given of the factor 2, starting in the most elementary\nconsiderations of geometry and kinematics of uniform acceleration, and moving\nto relativity, quantum mechanics and particle physics. The basic argument is\nthat in all the significant cases in which the factor 2 or 1/2 occurs in\nfundamental physics, whether classical, quantum or relativistic, the same\nphysical operation is taking place.\n", " Although cosmic expansion at very small distances is usually dismissed as\nentirely inconsequential, it appears that these extraordinarily small effects\nmay in fact have a real and significant influence on our world. Calculations\nsuggest that the minute recessional velocities associated with regions\nencompassed by extended bodies may have a role in creating the distinction\nbetween quantum and classical behavior. Using an uncertainty in position\nestimated from the spread in velocities associated with its size, the criterion\nthat the uncertainty in position should be smaller than the extension of the\nobject leads to a threshold size that could provide a fundamental limit\ndistinguishing the realm of objects governed by classical laws from those\ngoverned by quantum mechanics.\n", " Statistical classical mechanics and quantum mechanics are developed and\nwell-known theories that represent a basis for modern physics. The two\ndescribed theories are well known and have been well studied. As these theories\ncontain numerous paradoxes, many scientists doubt their internal consistencies.\nHowever, these paradoxes can be resolved within the framework of the existing\nphysics without the introduction of new laws. To clarify the paper for the\ninexperienced reader, we include certain necessary basic concepts of\nstatistical physics and quantum mechanics in this paper without the use of\nformulas. Exact formulas and explanations are included in the Appendices. The\ntext is supplemented by illustrations to enhance the understanding of the\npaper. The paradoxes underlying thermodynamics and quantum mechanics are also\ndiscussed. The approaches to the solutions of these paradoxes are suggested.\nThe first approach is dependent on the influence of the external observer\n(environment), which disrupts the correlations in the system. The second\napproach is based on the limits of the self-knowledge of the system for the\ncase in which both the external observer and the environment are included in\nthe considered system. The concepts of observable dynamics, ideal dynamics, and\nunpredictable dynamics are introduced. The phenomenon of complex (living)\nsystems is contemplated from the point of view of these dynamics.\n", " Following the spirit of de Broglie and Einstein, we think the concepts of\nmatter and radiation can be unified. We know a particle propagates like a wave;\nits motion is described by certain wave equations. At this point, it is not\nclear what the wave function represents. Besides the statistical meaning\nsuggested by the Copenhagen interpretation, does the wave function represent\nany physical motion? For photon, we know it is an electro-magnetic wave. But\nwhat about particles with rest mass, such as an electron? To investigate the\nphysical nature of matter wave, we propose that: (1) Like the photon, a\nparticle is an excitation wave of a real physical field. (2) Different types of\nparticles are different excitation modes of the same field. Based on this\nthinking, we show that the concept of quantum mechanics can be a natural\nextension of classical mechanics. By critically analyzing the transition from\nclassical physics to quantum physics, we found a new physical meaning for the\nquantum wave function. This work suggests that various quantum wave equations,\nincluding the Klein-Gordon equation, the Dirac equation and the Schrodinger\nequation, could have a common base relating to the Basic Wave Equation of the\nmatter wave. This work has some interesting implications. It suggests a\npossible way to explain the origin of visible matters and dark matters in our\nuniverse.\n", " In the present article we consider the inverse fallacy, a well known\ncognitive heuristic experimentally tested in cognitive science, which occurs\nfor intuitive judgments in situations of bounded rationality. We show that the\nquantum formalism can be used to describe in a very simple and general way this\nfallacy within the quantum formalism. Thus we suggest that in cognitive science\nthe formalism of quantum mechanics can be used to describe a quantum regime,\nthe bounded-rationality regime, where the cognitive heuristics are valid.\n", " According to special relativity and the equivalence principle, the Newtonian\ngravitational force between two particles with relativistic velocities\nincreases significantly with velocity and in fact becomes unbound as the latter\napproaches c. One may thus construct a deterministic model of three fast\nrotating light particles which uses gravitation as the attractive force,\nspecial relativity, the weak equivalence principle and introduces angular\nmomentum quantization as in the Bohr model of the H atom. The model shows the\nexistence of stable rotational states corresponding to highly relativistic\nparticle velocities with radii in the fm range. When the rest masses of the\nthree light particles is in the mass range of neutrinos (~0.1 eV/c^2), then\nsurprisingly the rest masses and Compton wavelengths of these rotational states\nare found to correspond to those of baryons and in fact the masses (~1 GeV/c^2)\nreproduce quite well the light baryon mass spectrum. Interestingly the model\nalso predicts a lower limit for the lifetime of the baryons which is consistent\nwith their measured lifetimes and yields a good approximation for their\nmagnetic moments. It thus appears that this deterministic Bohr-type approach to\nsubatomic phenomena, i.e. a classical special relativistic treatment coupled\nwith the de Broglie wavelength expression to seek conformity with quantum\nmechanics, may be a useful zeroth-order model to explore some quantum gravity\nproblems.\n", " The fundamental principle of quantum mechanics is that the probabilities of\nphysical outcomes are obtained from the intermediate states and processes of\nthe interacting particles, considered as happening concurrently. When the\ninteraction is described by a potential, the total energy of the particle is\nequal to its total kinetic plus potential energies. We derive the Schrodinger\nand Dirac equations as the conditions the wavefunction must satisfy at each\npoint in order to fulfill the corresponding energy equation. In our approach\nquantum theory is tangible, experimentally justified and theoretically\nconsistent. PACS numbers: 03.65.-w Keywords: Quantum principles; Quantum\ndynamics; Schrodinger equation; Dirac equation\n", " Modern undergraduate textbooks in electricity and magnetism typically focus\non a force representation of electrodynamics with an emphasis on Maxwell's\nEquations and the Lorentz Force Law. The vector potential $\\mathbf{A}$ and\nscalar potential $\\Phi$ play a secondary role mainly as quantities used to\ncalculate the electric and magnetic fields. However, quantum mechanics\nincluding quantum electrodynamics (QED) and other gauge theories demands a\npotential ($\\Phi$,$\\mathbf{A}$) oriented representation where the potentials\nare the more fundamental quantities. Here, we help bridge that gap by showing\nthat the homogeneous Maxwell's equations together with the Lorentz Force Law\ncan be derived from assuming that the potentials represent potential energy and\nmomentum per unit charge. Furthermore, we enumerate the additional assumptions\nthat are needed to derive the inhomogeneous Maxwell's equations. As part of\nthis work we demonstrate the physical nature and importance of gauge\ninvariance.\n", " In this paper we find a connection between the macroscopic classical laws of\ngases and the quantum mechanical description of molecules, composing an ideal\ngas. In such a gas, the motion of each individual molecule can be considered\nindependently on all other molecules, and thus the macroscopic parameters of\nideal gas, like pressure P and temperature T, can be introduced as a result of\nsimple averaging over all individual motions of molecules. It is shown that for\nan ideal gas enclosed in a macroscopic cubic box of volume V, the constant, in\nthe classical law of adiabatic expansion, i.e.PV^5/3=const, can be derived,\nbased on quantum mechanics. Physical implications of the result we disclose are\ndiscussed. In any case, our finding proves, seemingly for the first time, a\nmacroscopic manifestation of a quantum mechanical behavior, and this in\nrelation to classical thermodynamics.\n", " In this paper we consider the well-known cyclical model for climate change,\nfor example from glaciation to the present day and back, and compare this with\na two-state model in Quantum Mechanics. We establish the full correspondence.\n", " This is the reply to the paper of Andrei Khrennikov arXiv:1505.04293 in which\nhe expresses dissatisfaction with that the rough data in quantum experiments is\nnot easily available and compares it with the open rough data in genetics. I\ntry to explain why quantum experiments rough data is closed and why it differs\nradically from the biological. I also tried to answer the more thorny issue: is\nit possible to check quantum mechanics independently of other humans, e.g.\ntrusting nobody.\n", " The well known and oft-quoted Feynman's expression, entered the title,\nleading at a loss and even being objectionable, has not yet a clear\nexplanation. The hidden parameters problem in quantum mechanics is considered\nhere on the base of group-theoretic approach which includes the complete set of\nobservables indispensably. The last ones are the bilinear Hermitian forms\nconstructed from the Schroedinger equation solutions and its first derivatives,\nthey satisfy the algebraic completeness condition. These Hermitian forms,\nobtained for the simplest standard problem of particle transmission above\npotential step, had been compared with the Hermitian forms which are usually\nconsidered in this problem, and an additional ones, which may be obtained\nwithin the framework of an ordinary schemes of quantum mechanics. It is shown\nthat the generally recognised schemes of the problem solution lead to violation\nof some conservation laws on the step directly. On the contrary, the\ngroup-theoretic approach leads to fulfillments of all necessary conservation\nlaws everywhere at the same time. It is also shown that the complete set of\nobservables leads a probabilistic interpretation in quantum mechanics to be\nexcessive.\n", " Bernstein polynomials and B\\'ezier curves play an important role in\ncomputer-aided geometric design and numerical analysis, and their study relates\nto mathematical fields such as abstract algebra, algebraic geometry and\nprobability theory. We describe a theoretical framework that incorporates the\ndifferent aspects of the Bernstein-B\\'ezier theory, based on concepts from\ntheoretical physics. We relate B\\'ezier curves to the theory of angular\nmomentum in both classical and quantum mechanics, and describe physical\nanalogues of various properties of B\\'ezier curves -- such as their connection\nwith polar forms -- in the context of quantum spin systems. This previously\nunexplored relationship between geometric design and theoretical physics is\nestablished using the mathematical theory of Hamiltonian mechanics and\ngeometric quantization. An alternative description of spin systems in terms of\nharmonic oscillators serves as a physical analogue of P\\'olya's urn models for\nB\\'ezier curves. We relate harmonic oscillators to Poisson curves and the\nanalytical blossom as well. We present an overview of the relevant mathematical\nand physical concepts, and discuss opportunities for further research.\n", " In Part one of this Paper a hypothesis is forwarded of the electron charge in\nan atom existing in a distributed form. To check it by methods of\nelectrodynamics and mechanics (without invoking the formalism of quantum\nmechanics and the concepts of the wave function and of the operators), the\npotential, kinetic, and total energies were calculated for three states of the\nhydrogen atom, which were found to agree closely with the available\nexperimental data. The Part two of the Paper offers additional assumptions\nconcerning various scenarios of motion of elements of the distributed electron\ncharge which obey fully the laws of theoretical mechanics. The angular momentum\nof the ground-state hydrogen atom calculated in the frame of theoretical\nmechanics is shown to coincide with the spin which is $\\hbar/2$.\n", " The Ermakov equation, appearing in quantum mechanics of a harmonic\noscillator, is extended via dissipative and thermal terms to take into account\nthe effect of an environment.\n", " In the last twenty years modulation spaces, introduced by H. G. Feichtinger\nin 1983, have been successfully addressed to the study of signal analysis,\nPDE's, pseudodifferential operators, quantum mechanics, by hundreds of\ncontributions. In 2011 M. de Gosson showed that the time-frequency\nrepresentation Short-time Fourier Transform (STFT), which is the tool to define\nmodulation spaces, can be replaced by the Wigner distribution. This idea was\nfurther generalized to $\\tau$-Wigner representations in [9]. In this paper\ntime-frequency representations are viewed as images of symplectic matrices via\nmetaplectic operators. This new perspective highlights that the protagonists of\ntime-frequency analysis are metaplectic operators and symplectic matrices\n$\\mathcal{A} \\in Sp(2d,\\mathbb{R})$. We find conditions on $\\mathcal{A}$ for\nwhich the related symplectic time-frequency representation $W_\\mathcal{A}$ can\nreplace the STFT and give equivalent norms for weighted modulation spaces. In\nparticular, we study the case of covariant matrices $\\mathcal{A}$, i.e., their\ncorresponding $W_\\mathcal{A}$ are members of the Cohen class.\n Finally, we show that symplectic time-frequency representations\n$W_\\mathcal{A}$ can be efficiently employed in the study of Schr\\\"{o}dinger\nequations. This new approach may have further applications in quantum mechanics\nand PDE's.\n", " In this paper we study the dissipative effects and decoherence induced on a\nparticle moving at constant speed in front of a dielectric plate in quantum\nvacuum, developing a Closed-Time-Path (CTP) integral formulation in order to\naccount for the corrections to these phenomena generated by finite\ntemperatures. We compute the frictional force of the moving particle and find\nthat it contains two different contributions: a pure quantum term due to\nquantum fluctuations (even present at vanishing temperatures) and a temperature\ndependent component generated by thermal fluctuations (bigger contribution the\nhigher the temperature). We further estimate the decoherence timescale for the\ninternal degree of freedom of the quantum particle. As expected, decoherence\ntime is reduced by temperature, however, this feature is stronger for large\nvelocities and for resonant situations. When the particle approaches\nrelativistic speed, decoherence time becomes independent of temperature. The\nfinite temperatures corrections to the force or even in the decoherence\ntimescale could be used to track traces of quantum friction through the study\nof the velocity dependence since the solely evidence of this dependence\nprovides an indirect testimony of the existence of a quantum frictional force\n", " We analyze the logical foundations of quantum mechanics (QM) by stressing\nnon-objectivity of quantum observables which is a consequence of the absence of\nlogical atoms in QM. We argue that the matter of quantum non-objectivity is\nthat, on the one hand, the formalism of QM constructed as a mathematical theory\nis self-consistent, but, on the other hand, quantum phenomena as results of\nexperimenter's performances are not self-consistent. This self-inconsistency is\nan effect of that the language of QM differs much from the language of human\nperformances. The first is the language of a mathematical theory which uses\nsome Aristotelian and Russellian assumptions (e.g., the assumption that there\nare logical atoms). The second language consists of performative propositions\nwhich are self-inconsistent only from the viewpoint of conventional\nmathematical theory, but they satisfy another logic which is non-Aristotelian.\nHence, the representation of quantum reality in linguistic terms may be\ndifferent: from a mathematical theory to a logic of performative propositions.\nTo solve quantum self-inconsistency, we apply the formalism of non-classical\nself-referent logics.\n", " An interpretation of non-relativistic quantum mechanics is presented in the\nspirit of Erwin Madelung's hydrodynamic formulation of QM and Louis de\nBroglie's and David Bohm's pilot wave models. The aims of the approach are as\nfollows: 1) to have a clear ontology for QM, 2) to describe QM in a causal way,\n3) to get rid of the wave-particle dualism in pilot wave theories, 4) to\nprovide a theoretical framework for describing creation and annihilation of\nparticles, and 5) to provide a possible connection between particle QM and\nvirtual particles in QFT. These goals are achieved, if the wave function is\nreplaced by a fluid of so called virtual particles. It is also assumed that in\nthis fluid of virtual particles exist a few real particles and that only these\nreal particles can be directly observed. This has relevance for the measurement\nproblem in QM and it is found that quantum probabilities arise in a very\nnatural way from the structure of the theory. The model presented here is very\nsimilar to a recent computational model of quantum physics and recent Bohmian\nmodels of QFT.\n", " As it follows from the classical analysis of the data, the paper is intended\nto solve one of the main problems in the physics of tachyons, they are having\nimaginary mass. Further we use quantum mechanics to deal with the matters\nconcerning the particle. The paper also considers the physical viability of\ntachyons in real world. The paper also considers the natural occurrence and\npersistence of tachyons\n", " A general methodology and specific formalism are used to restrict the\nCopenhagen interpretation of quantum mechanics. A natural psi-collapse to\nreality is developed in an equation with terms independent of the measuring\nequipment.\n The theory requires one of three experiments for the ensemble average of\nposition, momentum, or energy, or a probability-experiment.\n Both classically and quantum mechanically, we define Artscience (AS) as\nLogic-Epistemology in Physics, where epistemology is treated as phenomenology.\nAS (logic phenomenology) is thus shown to be related to Physics\n(formalism-data) .\n", " Standard lore holds that magnetic forces are incapable of doing mechanical\nwork. More precisely, the claim is that whenever it appears that a magnetic\nforce is doing work, the work is actually being done by another force, with the\nmagnetic force serving only as an indirect mediator. However, the most familiar\ninstances of magnetic forces acting in everyday life, such as when bar magnets\nlift other bar magnets, appear to present manifest evidence of magnetic forces\ndoing work. These sorts of counterexamples are often dismissed as arising from\nquantum effects that lie outside the classical regime. In this paper, we show\nthat quantum theory is not needed to account for these phenomena, and that\nclassical electromagnetism admits a model of elementary magnetic dipoles on\nwhich magnetic forces can indeed do work. In order to develop this model, we\nrevisit the foundational principles of the classical theory of\nelectromagnetism, showcase the importance of constraints from relativity,\nexamine the structure of the multipole expansion, and study the connection\nbetween the Lorentz force law and conservation of energy and momentum.\n", " In a recent paper[1] Kundu and one of the present authors showed that there\nwere transient but observable CP violating effects in the decay of classical\ncurrents on the horizon of a black hole, if the Lagrangian of the Maxwell field\ncontained a CP violating angle {\\theta}. In this paper we demonstrate that a\nsimilar effect can be seen in the quantum mechanics of QED: a non-trivial Berry\nphase in the QED wave function is produced by in-falling electric charges. We\nalso investigate whether CP violation, of this or any other type, might be used\nto produce the baryon asymmetry of the universe, in models where primordial\nblack hole decay contributes to the matter content of the present universe.\nThis can happen both in a variety of hybrid inflation models, and in the\nHolographic Space-time (HST) model of inflation[2].\n", " It has for ages been a rather constant feature of thinking in science to take\nit for granted that the respective thinking happens in realms which are totally\noutside and independent of all the other phenomena that constitute the objects\nof such thinking. The imposition of this divide on two levels may conflict with\nbasic assumptions of Newtonian and Einsteinian mechanics, as well as with those\nin Quantum Mechanics. It also raises the question whether the realms in which\nthinking happens have no any other connection with the realms science deals\nwith, except to host and allow scientific thinking.\n", " We formulate the 2-body problem of electrodynamics using functional\ndifferential equations, and explain the peculiar features of these equations\nwhich indicate a paradigm shift in physics. We examine the possible empirical\nexistence of advanced interactions using various absorber theories of\nradiation. Functional differential equations were connected to quantum\nmechanics in a subsequent paper in this series and in later publications.\n", " Accurate models for open quantum systems -- quantum states that have\nnon-trivial interactions with their environment -- may aid in the advancement\nof a diverse array of fields, including quantum computation, informatics, and\nthe prediction of static and dynamic molecular properties. In recent years,\nquantum algorithms have been leveraged for the computation of open quantum\nsystems as the predicted quantum advantage of quantum devices over classical\nones may allow previously inaccessible applications. Accomplishing this goal\nwill require input and expertise from different research perspectives, as well\nas the training of a diverse quantum workforce, making a compilation of current\nquantum methods for treating open quantum systems both useful and timely. In\nthis Review, we first provide a succinct summary of the fundamental theory of\nopen quantum systems and then delve into a discussion on recent quantum\nalgorithms. We conclude with a discussion of pertinent applications,\ndemonstrating the applicability of this field to realistic chemical,\nbiological, and material systems.\n", " It is revisited the Michelson-Morley experiment within the quantum mechanics\nframework. One can define the wave function of photon in the whole space at a\ngiven moment of time. The phase difference between the source and receiver is a\ndistance between the source and receiver at the time of reception hence it does\nnot depend on the velocity of the frame. Then one can explain the null result\nof the Michelson-Morley experiment within the quantum mechanics framework.\n", " A wave function can be written in the form of {\\psi} = ReiS/h. We put this\nform of wave function into quantum mechanics equations and take hydrodynamic\nlimit, i. e., let Planck constant be zero. Then equations of motion (EOM)\ndescribing the movement of macroscopic bodies are retrieved. From Schrodinger\nequation, we obtain Newtonian mechanics, including Newtons three laws of\nmotion; from decouple Klein-Gordon equation with positive kinetic energy (PKE),\nwe obtain EOM of special relativity in classical mechanics. These are for PKE\nsystems. From negative kinetic energy (NKE) Schrodinger equation and decoupled\nKlein-Gordon equation, the EOM describing low momentum and relativistic motions\nof macroscopic dark bodies are derived. These are NKE systems, i. e., dark\nsystems. In all cases scalar and vector potentials are also taken into account.\nThe formalism obtained is collectively called macromechanics. For an isolated\nsystem containing PKE and NKE bodies, both total momentum and total kinetic\nenergy are conserved. A dark ideal gas produces a negative pressure, and its\nmicroscopic mechanism is disclosed. Two-body problems, where at least one is of\nNKE, are investigated for both macroscopic bodies and microscopic particles. A\nNKE proton and a PKE electron can compose a stable PKE atom, and its spectral\nlines have blue shifts compared to a hydrogen atom. The author suggests to seek\nfor these spectral lines in celestial spectra. This provides a way to seek for\ndark particles in space. Elastic collisions between a body and a dark body are\nresearched.\n", " We have studied optical metrics via null geodesics and optical-mechanical\nformulation of classical mechanics, and described the geometry and optics of\nmechanical systems with drag dependent quadratically on velocity. Then we\nstudied null geodesics as a central force system, deduced the related Binet's\nequation applied the analysis to other solutions of Einstein's equations in\nspherically symmetric spaces, paying special attention to the Tomimatsu-Sato\nmetric. Finally, we examined the dualities between different systems arising\nfrom conformal transformations that preserve the Jacobi metric.\n", " The Legendre transform is an important tool in theoretical physics, playing a\ncritical role in classical mechanics, statistical mechanics, and\nthermodynamics. Yet, in typical undergraduate or graduate courses, the power of\nmotivation and elegance of the method are often missing, unlike the treatments\nfrequently enjoyed by Fourier transforms. We review and modify the presentation\nof Legendre transforms in a way that explicates the formal mathematics,\nresulting in manifestly symmetric equations, thereby clarifying the structure\nof the transform algebraically and geometrically. Then we bring in the physics\nto motivate the transform as a way of choosing independent variables that are\nmore easily controlled. We demonstrate how the Legendre transform arises\nnaturally from statistical mechanics and show how the use of dimensionless\nthermodynamic potentials leads to more natural and symmetric relations.\n", " Postulates which lead to Minkowski spacetime are amended in a subtle way, and\nused to construct a consistent flat spacetime geometry with intrinsic quantum\ncharacter. Events in the new quantum geometry are described by labels of the\nform (t, x, s) where t (time) and x (spatial extent) are familiar and s is a\npositive integer. Structure-determining parameters for the novel spacetime are\nexplored in some detail. Particular structure parameters can be shown to\ngenerate equations of mainstream quantum mechanics, and structure parameters\ncan also be chosen which lead to the appearance of quantum particles as\nspatially extended moieties, such as strings. Further elaboration is required\nbefore the new spacetime can be used as a comprehensive basis for physical\ntheory, however even at the unfinished level a new way of accounting for rest\nmass emerges, and it is predicted that future experiments will fail to find\nHiggs particles.\n", " A distribution of electromagnetic fields presents a statistical assembly of a\nparticular type, which is at scale h a quantum statistical assembly itself and\nhas also been instrumental to concretisation of the basic probability\nassumption of quantum mechanics. Of specific concern in this discussion is an\nextensive train of radiation fields, of a total wave function \\psi, which are\ncontinuously (re)emitted and (re)absorbed by an oscillatory (point) charge of a\nzero rest mass and yet a finite dynamical mass, with the waves and charge\ntogether making up an extensive undulatory IED particle. The IED particle will\nas any real particle be subject to interactions with the environmental fields\nand particles, hence to excitations, and therefore will explore all possible\nstates over time; at scale $h$ the states are discrete. On the basis of the\nprinciples of statistics and statistical mechanics combined with first\nprinciples solutions for the IED particle, we derive for the IED particle the\nprobability functions in position space, of a form |\\psi|^2, and in\ndynamical-variable space.\n", " In this paper we prove the following: (1) The basic error of time-dependent\nperturbation theory is using the sum of first finite order of perturbed\nsolutions to substitute the exact solution in the divergent interval of the\nseries for calculating the transition probability. In addition quantum\nmechanics neglects the influence of the normality condition in the continuous\ncase. In both cases Fermi golden rule is not a mathematically reasonable\ndeductive inference from the Schrodinger equation. (2) The transition\nprobability per unit time deduced from the exact solution of the Schrodinger\nequation is zero, which cannot be used to describe the transition processes.\n", " The epistemological interpretation of quantum mechanics is still in an\nunacceptable status. This becomes obvious if looking on the variety of\ninterpretations currently under discussion. However, the physical community\ntogether with philosophers seem not to be willing to really face the problem\nand to find an explanation for the oddness of the foundation of nature, as the\nmathematical treatment of quantum physics can be judged as finished and perfect\nfor the use. The epistemology of quantum physics is often treated as a\nnegligible side effect, not worth to put effort in, as any outcome will not\naffect the calculation results at all. For the sake of human curiosity and to\nget a better understanding of the universe we are living in, this attitude\nshould be changed. A possibly interesting new approach is outlined in this\narticle, which may help to adjust the focus.\n", " There are different meanings of the term \"compositionality\" within science:\nwhat one researcher would call compositional, is not at all compositional for\nanother researcher. The most established conception is usually attributed to\nFrege, and is characterised by a bottom-up flow of meanings: the meaning of the\nwhole can be derived from the meanings of the parts, and how these parts are\nstructured together.\n Inspired by work on compositionality in quantum theory, and categorical\nquantum mechanics in particular, we propose the notions of Schrodinger,\nWhitehead, and complete compositionality. Accounting for recent important\ndevelopments in quantum technology and artificial intelligence, these do not\nhave the bottom-up meaning flow as part of their definitions.\n Schrodinger compositionality accommodates quantum theory, and also\nmeaning-as-context. Complete compositionality further strengthens Schrodinger\ncompositionality in order to single out theories like ZX-calculus, that are\ncomplete with regard to the intended model. All together, our new notions aim\nto capture the fact that compositionality is at its best when it is `real',\n`non-trivial', and even more when it also is `complete'.\n At this point we only put forward the intuitive and/or restricted formal\ndefinitions, and leave a fully comprehensive definition to future collaborative\nwork.\n", " General Theory of Relativity and Quantum theory gives two different\ndescription of the same mother nature in the big and small scale respectively.\nMathematical languages of these two theories are entirely different, one is\ngeometric while the other one is probabilistic. This curious feature makes the\nmerging of these theories (quantum-gravity) considerably difficult. In this\nmanuscript, we explore quantum-gravity in a dynamical system perspective. For\nexample, in the standard quantum theory, the wave equation is a linear partial\ndifferential equation while in General Theory of relativity, the field\nequations are highly nonlinear. A classical dynamical system can show very rich\nphase-space structures which is absent in a linear partial differential\nequation. In order to incorporate gravitational corrections one can think about\nthe nonlinear extensions of quantum theory. In this manuscript such an attempt\nto quantum gravity is explored. Dissipative and forcing corrections are found\nin the newly formulated quantum equation and it's physical interpretations are\ngiven.\n", " In a recent Foundations of Physics paper [5] by the current author it was\nshown that, when the self-force problem of classical electrodynamics is\nproperly solved, it becomes a plausible ontology underlying the statistical\ndescription of quantum mechanics. In the current paper we extend this result,\nshowing that ordinary matter, thus represented, possibly suffices in explaining\nthe outstanding observations currently requiring for this task the contrived\nnotions of dark-matter, dark-energy and inflation. The single mandatory `fix'\nto classical electrodynamics, demystifying both very small and very large scale\nphysics, should be contrasted with other ad hoc solutions to either problems.\nInstrumental to our cosmological model is scale covariance (and `spontaneous\nbreaking' thereof), a formal symmetry of classical electrodynamics treated on\nequal footing with its Poincare covariance, which is incompatible with the\n(absolute) metrical attributes of the GR metric tensor.\n", " The interferometry-based experimental tests of quantum properties of\nspace-time which the author sketched out in a recent short Letter [Nature 398\n(1999) 216] are here discussed in self-contained fashion. Besides providing\ndetailed derivations of the results already announced in the previous Letter,\nsome new results are also derived; in particular, the analysis is extended to a\nlarger class of scenarios for space-time fuzziness and an absolute bound on the\nmeasurability of the amplitude of a gravity wave is obtained. It is argued that\nthese studies could be helpful for the search of a theory describing a first\nstage of partial unification of Gravity and Quantum Mechanics.\n", " One of the oldest science problems - possibility of the perpetuum mobile is\ndiscussed. The interest to this problem was provoke a result, published in J.\nLow Temp.Phys. 112, 227 (1998); cond-mat/9811148, which contradicts to the\nsecond law of thermodynamics. According to this result, the thermal\nfluctuations can induce a voltage with direct component in a inhomogeneous\nsuperconducting ring at an unaltered temperature corresponded to the resistive\ntransition of the ring segment with the lowest critical temperature. This\nresult arises from obvious statements: 1) the switching of a ring segment\n$l_{b}$ into and out of the normal state, while the rest of the ring (segment\n$l_{a}$) remains superconducting, can induced a voltage with dc component (It\nis shown that, in spite of the wide spread opinion, this statement is correct\nbecause the superconductivity is a macroscopic quantum phenomena); 2) the\nthermal fluctuations switch the mesoscopic ring segment $l_{b}$ with lowest\ncritical temperature $T_{sb}$ into and out of the normal state at $T \\simeq\nT_{sb}$, while the rest of the ring remains superconducting if $T_{sa} > T\n\\simeq T_{sb}$. In order to resolve the contradiction between these obvious\nstatements and the second law of thermodynamics a possibility of the second\norder perpetuum mobile is considered theoretically. It is shown that from two\ntype of the perpetuum mobile, only type \"b\" and only in quantum systems is\npossible. According to the presented interpretation, the total entropy, as the\nmeasure of the chaos, may be systematically reduced in some quantum system\nbecause a \"switching\" between the classical and quantum mechanics is possible.\nInstruction for the making of the perpetuum mobile is enclosed.\n", " Action at distance in Newtonian physics is replaced by finite propagation\nspeeds in classical post--Newtonian physics. As a result, the differential\nequations of motion in Newtonian physics are replaced by functional\ndifferential equations, where the delay associated with the finite propagation\nspeed is taken into account. Newtonian equations of motion, with\npost--Newtonian corrections, are often used to approximate the functional\ndifferential equations. Are the finite propagation speeds the origin of the\nquantum mechanics? In this work a simple atomic model based on a functional\ndifferential equation which reproduces the quantized Bohr atomic model is\npresented. As straightforward application of the result the fine structure of\nthe hydrogen atom is approached.\n" ]

[ " This research article compares two characteristics of solutions of linear\ndifferential equations of the third order with variable coefficients. It\nappears that there is a lower estimate for wandering rate of solution to a\nlinear equation in terms of its frequency.\n", " Solvation is a notoriously difficult and nagging problem for the rigorous\ntheoretical description of chemistry in the liquid phase. Successes and\nfailures of various approaches ranging from implicit solvation modeling through\ndielectric continuum embedding and microsolvated quantum chemical modeling to\nexplicit molecular dynamics highlight this situation. Here, we focus on quantum\nchemical microsolvation and discuss an explicit conformational sampling ansatz\nto make this approach systematic. For this purpose, we introduce an algorithm\nfor the rolling and automated microsolvation of solutes. Our protocol takes\nconformational sampling and rearrangements in the solvent shell into account.\nIts reliability is assessed by monitoring the evolution of the spread and\naverage of the observables of interest.\n", " Hydrazine stands out as the most generally used chemical-reducing agent for\nreducing graphene oxide. Despite numerous experimental and theoretical\ninvestigations into the reduction reaction, the reduction mechanism remains\nunclear. In this study, we propose that, in aqueous hydrazine solutions, both\nhydrazine and hydroxide ions could initiate the reduction of graphene oxide. We\nintroduce a chemical reaction pathway involving C-H cleavage and a\ndehydroxylation process for the reduction of graphene oxide. By utilizing\ndensity functional theory calculations, the reduction reactions mediated by\nhydrazine and hydroxide ions are separately investigated. The reaction routes\non the basal plane and edge regions of graphene oxide are discussed\nindependently. The density functional theory calculations demonstrate that the\nproposed mechanism is both thermodynamically and dynamically feasible. This\nwork might contribute to an atomic-level comprehension of a longstanding\nchallenge in the field of graphene oxide.\n", " A comparison is made between conventional Michaelis-Menten kinetics and two\nmodels that take into account the duration of the conformational changes that\ntake place at the molecular level during the catalytic cycle of a monomer. The\nmodels consider the time that elapses from the moment an enzyme-substrate\ncomplex forms until the moment a product molecule is released, as well as the\nrecovery time needed to reset the conformational change that took place. In the\nfirst model the dynamics is described by a set of delayed differential\nequations, instead of the ordinary differential equations associated to\nMichaelis-Menten kinetics. In the second model the delay, the discretization\ninherent to enzyme reactions and the stochastic binding of substrates to\nenzimes at the molecular level is considered. All three models agree at\nequilibrium, as expected; however, out-of-equilibrium dynamics can differ\nsubstantially. In particular, both delayed models show oscillations at low\nvalues of the Michaelis constant which are not reproduced by the\nMichaelis-Menten model. Additionally, in certain cases, the dynamics shown by\nthe continuous delayed model differs from the dynamics of the discrete delayed\nmodel when some reactant become scarce.\n", " Copper nanopowder preparation and its X-Ray diffraction studies are reported\nin this paper. Electrolytic cathode deposition method is simple and cheapest\nprocess for its preparation. Copper nanopowder has been prepared from aqueous\ncopper sulphate solution. Wide range of experimental conditions has been\nadopted in this process and its X-Ray diffraction characterizations have been\nstudied. The results confirming copper nanopowder with size below 30 nm.\nUniformed size Copper nanopowder preparation, in normal room temperature is\nimportance of this study.\n", " In most natural sciences there is currently the insight that it is necessary\nto bridge gaps between different processes which can be observed on different\nscales. This is especially true in the field of chemical reactions where the\nabilities to form bonds between different types of atoms and molecules create\nmuch of the properties we experience in our everyday life, especially in all\nbiological activity. There are essentially two types of processes related to\nbiochemical reaction networks, the interactions among molecules and\ninteractions involving their conformational changes, so in a sense, their\ninternal state. The first type of processes can be conveniently approximated by\nthe so-called mass-action kinetics, but this is not necessarily so for the\nsecond kind where molecular states do not define any kind of density or\nconcentration. In this paper we demonstrate the necessity to study reaction\nnetworks in a stochastic formulation for which we can construct a coherent\napproximation in terms of specific space-time scales and the number of\nparticles. The continuum limit procedure naturally creates equations of\nFokker-Planck type where the evolution of the concentration occurs on a slower\ntime scale when compared to the evolution of the conformational changes, for\nexample triggered by binding or unbinding events with other (typically smaller)\nmolecules. We apply the asymptotic theory to derive the effective, i.e.\nmacroscopic dynamics of the biochemical reaction system. The theory can also be\napplied to other processes where entities can be described by finitely many\ninternal states, with changes of states occuring by arrival of other entities\ndescribed by a birth-death process.\n", " The kinematics laws of electrons motion in the volume of a planar vacuum\ndiode running in regime of current saturation are used. The characteristics of\ndiode hysteresis in the conditions of S-figurative instability are got and\nanalyzed.\n", " During the last decade, network approaches became a powerful tool to describe\nprotein structure and dynamics. Here we review the links between disordered\nproteins and the associated networks, and describe the consequences of local,\nmesoscopic and global network disorder on changes in protein structure and\ndynamics. We introduce a new classification of protein networks into\ncumulus-type, i.e., those similar to puffy (white) clouds, and stratus-type,\ni.e., those similar to flat, dense (dark) low-lying clouds, and relate these\nnetwork types to protein disorder dynamics and to differences in energy\ntransmission processes. In the first class, there is limited overlap between\nthe modules, which implies higher rigidity of the individual units; there the\nconformational changes can be described by an energy transfer mechanism. In the\nsecond class, the topology presents a compact structure with significant\noverlap between the modules; there the conformational changes can be described\nby multi-trajectories; that is, multiple highly populated pathways. We further\npropose that disordered protein regions evolved to help other protein segments\nreach rarely visited but functionally-related states. We also show the role of\ndisorder in spatial games of amino acids; highlight the effects of\nintrinsically disordered proteins (IDPs) on cellular networks and list some\npossible studies linking protein disorder and protein structure networks.\n", " Template directed replication of nucleic acids is at the essence of all\nliving beings and a major milestone for any origin of life scenario. We here\npresent an idealized model of prebiotic sequence replication, where binary\npolymers act as templates for their autocatalytic replication, thereby serving\nas each others reactants and products in an intertwined molecular ecology. Our\nmodel demonstrates how autocatalysis alters the qualitative and quantitative\nsystem dynamics in counter-intuitive ways. Most notably, numerical simulations\nreveal a very strong intrinsic selection mechanism that favours the appearance\nof a few population structures with highly ordered and repetitive sequence\npatterns when starting from a pool of monomers. We demonstrate both\nanalytically and through simulation how this \"selection of the dullest\" is\ncaused by continued symmetry breaking through random fluctuations in the\ntransient dynamics that are amplified by autocatalysis and eventually propagate\nto the population level. The impact of these observations on related prebiotic\nmathematical models is discussed.\n", " We investigated Capacitance-Voltage (C-V) characteristics of the Depletion\nMode Buried Channel InGaAs/InAs Quantum Well FET by using Self-Consistent\nmethod incorporating Quantum Mechanical (QM) effects. Though the experimental\nresults of C-V for enhancement type device is available in recent literature, a\ncomplete characterization of electrostatic property of depletion type Buried\nChannel Quantum Well FET (QWFET) structure is yet to be done. C-V\ncharacteristics of the device is studied with the variation of three important\nprocess parameters: Indium (In) composition, gate dielectric and oxide\nthickness. We observed that inversion capacitance and ballistic current tend to\nincrease with the increase in Indium (In) content in InGaAs barrier layer.\n", " In this paper we study the main characteristics of some evaluation codes\nparameterized by the edges of a bipartite graph with a perfect matching.\n", " Molecular processes of neuronal learning have been well-described. However,\nlearning mechanisms of non-neuronal cells have not been fully understood at the\nmolecular level. Here, we discuss molecular mechanisms of cellular learning,\nincluding conformational memory of intrinsically disordered proteins and\nprions, signaling cascades, protein translocation, RNAs (microRNA and lncRNA),\nand chromatin memory. We hypothesize that these processes constitute the\nlearning of signaling networks and correspond to a generalized Hebbian learning\nprocess of single, non-neuronal cells, and discuss how cellular learning may\nopen novel directions in drug design and inspire new artificial intelligence\nmethods.\n", " The main purposes of this paper are (i) to illustrate explicitly by a number\nof examples the gauge functions chi(x, t) whose spatial and temporal\nderivatives transform one set of electromagnetic potentials into another\nequivalent set; and (ii) to show that, whatever propagation or non-propagation\ncharacteristics are exhibited by the potentials in a particular gauge, the\nelectric and magnetic fields are always the same and display the experimentally\nverified properties of causality and propagation at the speed of light. The\nexample of the transformation from the Lorenz gauge (retarded solutions for\nboth scalar and vector potential) to the Coulomb gauge (instantaneous,\naction-at-a-distance, scalar potential) is treated in detail. A transparent\nexpression is obtained for the vector potential in the Coulomb gauge, with a\nfinite nonlocality in time replacing the expected spatial nonlocality of the\ntransverse current. A class of gauges (v-gauge) is described in which the\nscalar potential propagates at an arbitrary speed v relative to the speed of\nlight. The Lorenz and Coulomb gauges are special cases of the v-gauge. The last\nexamples of gauges and explicit gauge transformation functions are the\nHamiltonian or temporal gauge, the nonrelativistic Poincare or multipolar\ngauge, and the relativistic Fock-Schwinger gauge.\n", " We use deep neural networks (DNN) to obtain the microscopic characteristics\nof partons in terms of dynamical degrees of freedom on the basis of an\noff-shell quasiparticle description. We aim to infer masses and widths of\nquasi-gluons, up/down, and strange quarks using constraints on the macroscopic\nthermodynamic observables obtained by the first-principles calculations lattice\nQCD. In this work, we use 3 independent dimensionless thermodynamic observables\nfrom lQCD for minimization. First, we train our DNN using the DQPM (Dynamical\nQuasiParticle Model) Ansatz for the masses and widths. Furthermore, we use the\nDNN capabilities to generalize this Ansatz, to evaluate which quasiparticle\ncharacteristics are desirable to describe different thermodynamic functions\nsimultaneously. To evaluate consistently the microscopic properties obtained by\nthe DNN in the case of off-shell quarks and gluons, we compute transport\ncoefficients using the spectral function within Kubo-Zubarev formalism in\ndifferent setups. In particular, we make a comprehensive comparison in the case\nof the dimensionless ratios of shear viscosity over entropy density $\\eta/s$\nand electric conductivity over temperature $\\sigma_Q/T$, which provide\nadditional constraints for the parameter generalization of the considered\nmodels.\n", " Modelling the Calvin cycle of photosynthesis leads to various systems of\nordinary differential equations and reaction-diffusion equations. They differ\nby the choice of chemical substances included in the model, the choices of\nstoichiometric coefficients and chemical kinetics and whether or not diffusion\nis taken into account. This paper studies the long-time behaviour of solutions\nof several of these systems, concentrating on the ODE case. In some examples it\nis shown that there exist two positive stationary solutions. In several cases\nit is shown that there exist solutions where the concentrations of all\nsubstrates tend to zero at late times and others (runaway solutions) where the\nconcentrations of all substrates increase without limit. In another case, where\nthe concentration of ATP is explicitly included, runaway solutions are ruled\nout.\n", " We present norm estimates for solutions of first and second order elliptic\nBVPs of the Dirac operator considered over a bounded and smooth domain of the\nn-dimension Euclidean space. The solutions whose norms to be estimated are in\nsome Sobolev spaces and the boundary conditions as traces of solutions and\ntheir derivatives are in some Slobodeckji spaces.\n", " We use the framework of the first-order differential structure in metric\nmeasure spaces introduced by Gigli to define a notion of weak solutions to\ngradient flows of convex, lower semicontinuous and coercive functionals. We\nprove their existence and uniqueness and show that they are also variational\nsolutions; in particular, this is an existence result for variational\nsolutions. Then, we apply this technique in the case of a gradient flow of a\nfunctional with inhomogeneous growth.\n", " In this paper we use considerations of non-commutative geometry to deduce a\nmodel for QCD interactions. The model also explains within the same theoretical\nframework hitherto purely phenomenological characteristics of the quarks like\ntheir fractional charge, mass,handedness and confinement.\n", " The paper discusses the affection of the scalar potential on the speed of\nlight in the electromagnetic field, by means of the characteristics of\noctonion. In the octonion space, the radius vector is combined with the\nintegral of field potentials to become one new radius vector. When the field\npotentials can not be neglected, the new radius vector will cause the\nprediction to departure slightly from the theoretical value of the speed of\nlight. The results explain why the speed of light varies in diversiform optical\nwaveguide. And there exist negative refractive indexes due to different scalar\npotentials in the gravitational field and electromagnetic field.\n", " Protocells are supposed to have played a key role in the self-organizing\nprocesses leading to the emergence of life. Existing models either (i) describe\nprotocell architecture and dynamics, given the existence of sets of\ncollectively self-replicating molecules for granted, or (ii) describe the\nemergence of the aforementioned sets from an ensemble of random molecules in a\nsimple experimental setting (e.g. a closed system or a steady-state flow\nreactor) that does not properly describe a protocell. In this paper we present\na model that goes beyond these limitations by describing the dynamics of sets\nof replicating molecules within a lipid vesicle. We adopt the simplest possible\nprotocell architecture, by considering a semi-permeable membrane that selects\nthe molecular types that are allowed to enter or exit the protocell and by\nassuming that the reactions take place in the aqueous phase in the internal\ncompartment. As a first approximation, we ignore the protocell growth and\ndivision dynamics. The behavior of catalytic reaction networks is then\nsimulated by means of a stochastic model that accounts for the creation and the\nextinction of species and reactions. While this is not yet an exhaustive\nprotocell model, it already provides clues regarding some processes that are\nrelevant for understanding the conditions that can enable a population of\nprotocells to undergo evolution and selection.\n", " We present a particular class of solutions in Cartesian, cylindrical and\nspherical coordinates of the non-dispersive travelling wave variety that\npropagate an envelope of varying vorticity some of which include topological\nwaves with parallel electric and magnetic components. The significance of these\nsolutions is examined in the recently proposed Axion-Maxwell field theory with\npotential applications in material science and topological insulators.\n", " The possible transition of large mass magnetic monopole solutions to monopole\nblack hole solutions in Yang-Mills-Higgs theory with spontaneous symmetry\nbreaking (SSB) coupled to the low energy effective theory of superstring with\nonly dilatonic sector is studied. Our main motivation is to explore the effects\nof the dilaton field on the monopole black hole solutions. Working in the\nEinstein conformal gauge, it is explicitly shown, in terms of the Hawking\nevaporation of the monopole black hole, that the presence of the dilaton field\nappears to introduce a \"mass dependent extra attractive interaction\" in\naddition to that of the standard Einstein gravity into the system.\n", " Based on the electromagnetic interaction and gravitational interaction, the\nquantum interplays of the ordinary matter and the dark matter in the octonion\nspace are discussed. The paper presents the quantization of the particles of\nordinary matter and dark matter, including the quantization of the\nelectromagnetic field and gravitational field etc. In the electromagnetic and\ngravitational interactions, it deduces some predictions of the field source\nparticle, which are consistent with the Dirac equation and Schrodinger equation\nin the quantum mechanics. The researches show that there exist some quantum\ncharacteristics of the intermediate particle in the electromagnetic and\ngravitational interactions, including the Dirac-like equation etc.\n", " Using the technique of the elliptic Frobenius determinant, we construct new\nelliptic solutions of the $QD$-algorithm. These solutions can be interpreted as\nelliptic solutions of the discrete-time Toda chain as well. As a by-product, we\nobtain new explicit orthogonal and biorthogonal polynomials in terms of the\nelliptic hypergeometric function ${_3}E_2(z)$. Their recurrence coefficients\nare expressed in terms of the elliptic functions. In the degenerate case we\nobtain the Krall-Jacobi polynomials and their biorthogonal analogs.\n", " Folding of proteins into their correct native structure is key to their\nfunction. Simultaneously, the intricate interplay between cell movement and\nprotein conformation highlights the complex nature of cellular processes. In\nthis work, we demonstrate the impact of Terahertz (THz) signaling on\ncontrolling protein conformational changes in a random medium. Our system of\ninterest consists of a communication link that involves a nanoantenna\ntransmitter, a protein receiver, and a channel composed of moving red blood\ncells. Due to the system dynamics, we investigate the influence of both the\nfast and slow channel variations on protein folding. Specifically, we analyze\nthe system's selectivity to asses the effectiveness of the induced THz\ninteraction in targeting a specific group of proteins under fading conditions.\nBy optimizing the selectivity metric with respect to the nanoantenna power and\nfrequency, it is possible to enhance the controllability of protein\ninteractions. Our probabilistic analysis provides a new perspective regarding\nelectromagnetically triggered protein molecules, their microenvironment and\ntheir interaction with surrounding particles. It helps elucidate how external\nconditions impact the protein folding kinetics and pathways. This results in\nnot only understanding the mechanisms underlying THz-induced protein\ninteractions but also engineering these still-emerging tools.\n", " Calmodulin, the ubiquitous calcium-activated second messenger in eukaryotes,\nis an extremely versatile molecule involved in many biological processes:\nmuscular contraction, synaptic plasticity, circadian rhythm, and cell cycle,\namong others. The protein is structurally organised into two globular lobes,\njoined by a flexible linker. Calcium modulates calmodulin activity by favoring\na conformational transition of each lobe from a closed conformation to an open\nconformation. Most targets have a strong preference for one conformation over\nthe other, and depending on the free calcium concentration in a cell,\nparticular sets of targets will preferentially interact with calmodulin. In\nturn, targets can increase or decrease the calcium affinity of the calmodulin\nmolecules to which they bind. Interestingly, experiments with the tryptic\nfragments showed that most targets have a much lower affinity for the N-lobe\nthan for the C-lobe. Hence, the latter predominates in the formation of most\ncalmodulin-target complexes. We showed that a relatively simple allosteric\nmechanism, based the classic MWC model, can capture the observed modulation of\nboth the isolated C-lobe, and intact calmodulin, by individual targets.\nMoreover, our model can be naturally extended to study how the calcium affinity\nof a single pool of calmodulin is modulated by a mixture of competing targets\nin vivo.\n", " Boolean networks have been widely used to explore aspects of gene regulation,\ntraditionally with a single network. A modified form of the model to explore\nthe effects of increasing the number of gene states has also recently been\nintroduced. In this paper, these discrete dynamical networks are evolved as\ndiploids within rugged fitness landscapes to explore their behaviour. Results\nsuggest the general properties of haploid networks in similar circumstances\nremain for diploids. The previously proposed inherent fitness landscape\nsmoothing properties of eukaryotic sex are shown to be exhibited in these\ndynamical systems, as is their propensity to change in size based upon the\ncharacteristics of the network and fitness landscape.\n", " In this paper, we explore static spherically symmetric wormhole solutions in\nthe framework of $n$-dimensional Einstein Gauss-Bonnet gravity. Our objective\nis to find out wormhole solutions that satisfy energy conditions. For this\npurpose, we consider two frameworks such as Gaussian distributed and Lorentzian\ndistributed non-commutative geometry. Taking into account constant redshift\nfunction, we obtain solutions in the form of shape function. The fifth and\nsixth dimensional solutions with positive as well as negative Gauss-Bonnet\ncoefficient are discussed. Also, we check the equilibrium condition for the\nwormhole solutions with the help of generalized Tolman-Oppenheimer-Volkov\nequation. It is interesting to mention here that we obtain fifth dimensional\nstable wormhole solutions in both distributions that satisfy the energy\nconditions.\n", " Power series expansions naturally arise whenever solutions of ordinary\ndifferential equations are studied in the regime of perturbation theory. In the\ncase of quasi-periodic solutions the issue of convergence of the series is\nplagued of the so-called small divisor problem. In this paper we review a\nmethod recently introduced to deal with such a problem, based on\nrenormalisation group ideas and multiscale techniques. Applications to both\nquasi-integrable Hamiltonian systems (KAM theory) and non-Hamiltonian\ndissipative systems are discussed. The method is also suited to situations in\nwhich the perturbation series diverges and a resummation procedure can be\nenvisaged, leading to a solution which is not analytic in the perturbation\nparameter: we consider explicitly examples of solutions which are only\ninfinitely differentiable in the perturbation parameter, or even defined on a\nCantor set.\n", " Stochastic kinetic models of genetic expression are able to describe protein\nfluctuations. A comparative study of the canonical and a feedback model is\ngiven here by using stochastic simulation methods. The feedback model is\nskeleton model implementation of the circular gene hypothesis, which suggests\nthe interaction between the synthesis and degradation of mRNA. Qualitative and\nquantitative changes in the shape and in the numerical characteristics of the\nstationary distributions suggest that more combined experimental and\ntheoretical studies should be done to uncover the details of the kinetic\nmechanism of gene expression.\n", " The methods of statistical physics of open systems are used for describing\nthe time dependence of economic characteristics (income, profit, cost, supply,\ncurrency etc.) and their correlations with each other. Nonlinear equations\n(analogies of known reaction-diffusion, kinetic, Langevin equation) describing\nappearance of bifurcations, self-sustained oscillational processes,\nself-organizations in economic phenomena are offered.\n", " Localization phenomena in light scattering from random fluctuations of matter\nfields and spacetime metric near a black hole horizon were predicted to produce\na pronounced peak in the angular distribution of second harmonic light in the\ndirection normal to the horizon. Therefore, detection of second harmonic\ngeneration may become a viable observational tool to study spacetime physics\nnear event horizons of astronomical black holes. The light localization\nphenomena near the horizon may be facilitated by the existence of surface\nelectromagnetic wave solutions. In this communication we study such surface\nelectromagnetic wave solutions near the horizon of a Schwarzschild metric\ndescribing a black hole in vacuum. We demonstrate that such surface wave\nsolutions must appear when quantum gravity effects are taken into account.\nPotential observational evidence of this effect are also discussed.\n", " Anomalous fermion number violation is studied in the background of a pure\nSU(2) gauge field in Minkowski space using the method of N. Christ. It is\ndemonstrated that the chiral fermion number is violated by at most an integer\namount. Then the method is applied for a spherically symmetric Minkowski space\nclassical gauge field in the background. These classical gauge fields are\nfinite energy solutions to pure SU(2) equations of motion with in general\nnon-integer topological charge. We show that in the classical background which\nduring a finite time-interval matches such solutions the fermion number\nviolation is integer and non-zero. In particular, we calculate the violation of\nthe fermion number in the presence of L\\\"uscher-Schechter solutions. The\nmeaning of anomaly equation and applications to QCD and electroweak theory are\nbriefly discussed. We also comment on the relation of the results of this paper\nto the previous work.\n", " We consider the thin-film equation for a class of free boundary conditions\nmodelling friction at the contact line, as introduced by E and Ren. Our\nanalysis focuses on formal long-time asymptotics of solutions in the perfect\nwetting regime. In particular, through the analysis of quasi-self-similar\nsolutions, we characterize the profile and the spreading rate of solutions\ndepending on the strength of friction at the contact line, as well as their\n(global or local) corrections, which are due to the dynamical nature of the\nfree boundary conditions. These results are complemented with full transient\nnumerical solutions of the free boundary problem.\n", " Molecular property prediction is a critical task in computational drug\ndiscovery. While recent advances in Graph Neural Networks (GNNs) and\nTransformers have shown to be effective and promising, they face the following\nlimitations: Transformer self-attention does not explicitly consider the\nunderlying molecule structure while GNN feature representation alone is not\nsufficient to capture granular and hidden interactions and characteristics that\ndistinguish similar molecules. To address these limitations, we propose SYN-\nFUSION, a novel approach that synergistically combines pre-trained features\nfrom GNNs and Transformers. This approach provides a comprehensive molecular\nrepresentation, capturing both the global molecule structure and the individual\natom characteristics. Experimental results on MoleculeNet benchmarks\ndemonstrate superior performance, surpassing previous models in 5 out of 7\nclassification datasets and 4 out of 6 regression datasets. The performance of\nSYN-FUSION has been compared with other Graph-Transformer models that have been\njointly trained using a combination of transformer and graph features, and it\nis found that our approach is on par with those models in terms of performance.\nExtensive analysis of the learned fusion model across aspects such as loss,\nlatent space, and weight distribution further validates the effectiveness of\nSYN-FUSION. Finally, an ablation study unequivocally demonstrates that the\nsynergy achieved by SYN-FUSION surpasses the performance of its individual\nmodel components and their ensemble, offering a substantial improvement in\npredicting molecular properties.\n", " The study on the linkage of elementary particle mass differences with\npion-muon mass difference is explored further. In the present study we show\nthis linkage to be equally true for the mass differences amongst the members of\nSU(3) hadron multiplets and the hadrons belonging to multiplets having\ndifferent spin and parity characteristics. This reinforces our contention that\ninter particle mass excitations are quantized as integral multiples of basic\nmass unit of 29.318 MeV and thereby indicates the fundamental nature of\npion-muon mass difference in the elementary particle mass distribution.\n", " We consider a linear size-structured population model with diffusion in the\nsize-space. Individuals are recruited into the population at arbitrary sizes.\nThe model is equipped with generalized Wentzell-Robin (or dynamic) boundary\nconditions. This allows modelling of \"adhesion\" at extremely small or large\nsizes. We establish existence and positivity of solutions by showing that\nsolutions are governed by a positive quasicontractive semigroup of linear\noperators on the biologically relevant state space. This is carried out via\nestablishing dissipativity of a suitably perturbed semigroup generator. We also\nshow that solutions of the model exhibit balanced exponential growth, that is\nour model admits a finite dimensional global attractor. In case of strictly\npositive fertility we are able to establish that solutions in fact exhibit\nasynchronous exponential growth.\n", " We prove a number of \\textit{a priori} estimates for weak solutions of\nelliptic equations or systems with vertically independent coefficients in the\nupper-half space. These estimates are designed towards applications to boundary\nvalue problems of Dirichlet and Neumann type in various topologies. We work in\nclasses of solutions which include the energy solutions. For those solutions,\nwe use a description using the first order systems satisfied by their conormal\ngradients and the theory of Hardy spaces associated with such systems but the\nmethod also allows us to design solutions which are not necessarily energy\nsolutions. We obtain precise comparisons between square functions,\nnon-tangential maximal functions and norms of boundary trace. The main thesis\nis that the range of exponents for such results is related to when those Hardy\nspaces (which could be abstract spaces) are identified to concrete spaces of\ntempered distributions. We consider some adapted non-tangential sharp functions\nand prove comparisons with square functions. We obtain boundedness results for\nlayer potentials, boundary behavior, in particular strong limits, which is new,\nand jump relations. One application is an extrapolation for solvability ''\\'a\nla {\\v{S}}ne{\\u\\i}berg\". Another one is stability of solvability in perturbing\nthe coefficients in $L^\\infty$ without further assumptions. We stress that our\nresults do not require De Giorgi-Nash assumptions, and we improve the available\nones when we do so.\n", " Bacterial sensor systems can be used for the detection and measurement of\nmolecular signal concentrations. The dynamics of the sensor directly depend on\nthe biological properties of the bacterial sensor cells; manipulation of these\nfeatures in the wet lab enables the engineering and optimization of the\nbacterial sensor kinetics. This necessitates the development of biologically\nmeaningful computational models for bacterial sensors comprising a variety of\ndifferent molecular mechanisms, which further facilitates a systematic and\nquantitative evaluation of optimization strategies. In this work, we dissect\nthe detection chain of bacterial sensors from a mathematical perspective from\nwhich we derive, supported by wet-lab data, a complete computational model for\na Streptococcus mutans-based bacterial sensor as a case example. We address the\nengineering of bacterial sensors by investigating the impact of altered\nbacterial cell properties on the sensor response characteristics, specifically\nsensor sensitivity and response signal intensity. This is achieved through a\nsensitivity analysis targeting both the steady-state and transient sensor\nresponse characteristics. Alongside the demonstration of suitability of our\nmethodological approach, our analysis shows that an increase of sensor\nsensitivity, through a targeted manipulation of bacterial physiology, often\ncomes at the cost of generally diminished sensor response intensity.\n", " A time-variant analogue of an interpolation problem equivalent to the relaxed\ncommutant lifting problem is introduced and studied. In a somewhat less general\nform the problem already appears in the analysis of the set of all solutions to\nthe three chain completion problem. The interpolants are upper triangular\noperator matrices of which the columns induce contractive operators. The set of\nall solutions of the problem is described explicitly. The results presented are\ntime-variant analogues of the main theorems in [A.E. Frazho, S. ter Horst, and\nM.A. Kaashoek, All solutions to the relaxed commutant lifting problem, Acta\nSci. Math. (Szeged) 72 (2006), 299--318].\n", " We study the effects of a static and uniform magnetic field on the evolution\nof energy density fluctuations present in a medium. By numerically solving the\nrelativistic Boltzmann-Vlasov equation within the relaxation time\napproximation, we explicitly show that magnetic field can affect the\ncharacteristics of energy density fluctuations at the timescale the system\nachieves local thermodynamic equilibrium. A detailed momentum mode analysis of\nfluctuations reveals that magnetic field increases the damping of mode\noscillations, especially for the low momentum modes. This leads to a reduction\nin the ultraviolet (high momentum) cutoff of fluctuations and also slows down\nthe dissipation of relatively low momentum fluctuation modes. We discuss the\nphenomenological implications of our study on various sources of fluctuations\nin relativistic heavy-ion collisions.\n", " We apply an analogue of the Zakharov-Shabat dressing method to obtain\ninfinite matrix solutions to the Toda lattice hierarchy. Using an operator\ntransformation we convert some of these into solutions in terms of integral\noperators and Fredholm determinants. Others are converted into a class of\noperator solutions to the $l$-periodic Toda hierarchy.\n", " We study the $\\epsilon \\to 0$ behavior of recurrence relations of the type\n$\\sum_{j=0}^l a_j(k\\epsilon,\\epsilon)y_{k+j}=0,$ $k\\in \\zdd$ ($l$ fixed). The\n$a_j$ are $C^{\\infty}$ functions in each variable on $I\\times [0,\\e_0]$ for a\nbounded interval $I$ and $\\e_0>0$. Under certain regularity assumptions we find\nthe asymptotic behavior of the solutions of such recurrences. In typical cases\nthere exists a fundamental set of solutions in the form $\\{\\exp(\\epi\nF_m(k\\epsilon,\\epsilon))\\}_{m=1... l}$ where the functions $F_m$ are\n$C^{\\infty}$ in each variable on the same domain as the $a_j$, showing in\nparticular that the formal perturbation-series solutions are asymptotic to true\nsolutions of these recurrences. Some applications are also briefly discussed.\n", " We show that based on the general solution, given by Corrigan, Olive, Fairlie\nand Nuyts, in the region outside the monopole's core; the equations of motion\nin the Higgs vacuum (i.e. outside the monopole's core) will not allow\nasymptotically non-singular extended non-trivial non-Dyonic (including, also,\nall static) solutions of the 't Hooft-Polyakov monopole. In other words, unless\nthe monopole's magnetic charge is shielded (by some mechanism), the Dirac\nstring is inevitable asymptotically, in the region outside the monopole's core,\nfor all non-Dyonic solutions that are admissible by the equations of motion.\nThat we show that the non-dyonic solutions (based on Corrigan et al) will\ninclude all \"admissible\" static solutions and their gauge transform might be\ninterpreted as that all admissible dyonic solutions (based on Corrigan et al)\nare composite solutions.\n", " In this paper, we establish a formula determining the value of the Cauchy\nintegrals of the real and purely imaginary Ablowitz-Segur solutions for the\ninhomogeneous second Painlev\\'e equation. Our approach relies on the analysis\nof the corresponding Riemann-Hilbert problem and the construction of an\nappropriate parametrix in a neighborhood of the origin. Obtained integral\nformulas are consistent with already known analogous results for Ablowitz-Segur\nsolutions of homogeneous Painlev\\'e II equation.\n", " Controlled assembly of single-crystal, colloidal maghemite nanoparticles is\nfacilitated via a high-temperature polyol-based pathway. Structural\ncharacterization shows that size-tunable nanoclusters of 50 and 86 nm diameters\n(D), with high dispersibility in aqueous media, are composed of $\\sim$ 13 nm\n(d) crystallographically oriented nanoparticles. The interaction effects are\nexamined against the increasing volume fraction, $\\phi$, of the inorganic\nmagnetic phase that goes from individual colloidal nanoparticles ($\\phi$= 0.47)\nto clusters ($\\phi$= 0.72). The frozen-liquid dispersions of the latter exhibit\nweak ferrimagnetic behavior at 300 K. Comparative Mossbauer spectroscopic\nstudies imply that intra-cluster interactions come into play. A new insight\nemerges from the clusters temperature-dependent ac susceptibility that displays\ntwo maxima in $\\chi$''(T), with strong frequency dispersion. Scaling-law\nanalysis, together with the observed memory effects suggest that a superspin\nglass state settles-in at T$_{B}$ $\\sim$ 160-200 K, while at\nlower-temperatures, surface spin-glass freezing is established at T$_{f}$\n$\\sim$40- 70 K. In such nanoparticle-assembled systems, with increased $\\phi$,\nMonte Carlo simulations corroborate the role of the inter-particle dipolar\ninteractions and that of the constituent nanoparticles surface spin disorder in\nthe emerging spin-glass dynamics.\n", " Nonuniformities in the imaging characteristics of modern image sensors are a\nprimary factor in the push to develop a pixel-level generalization of the\nphoton transfer characterization method. In this paper, we seek to develop a\nbody of theoretical results leading toward a comprehensive approach for\ntackling the biggest obstacle in the way of this goal: a means of pixel-level\nconversion gain estimation. This is accomplished by developing an estimator for\nthe reciprocal-difference of normal variances and then using this to construct\na novel estimator of the conversion gain. The first two moments of this\nestimator are derived and used to construct exact and approximate confidence\nintervals for its absolute relative bias and absolute coefficient of variation,\nrespectively. A means of approximating and computing optimal sample sizes are\nalso discussed and used to demonstrate the process of pixel-level conversion\ngain estimation for a real image sensor.\n", " A general formalism to solve nonlinear differential equations is given.\nSolutions are found and reduced to those of second order nonlinear differential\nequations in one variable. The approach is uniformized in the geometry and\nsolves generic nonlinear systems. Further properties characterized by the\ntopology and geometry of the associated manifolds may define global properties\nof the solutions.\n", " The $U(1)_A$ boson mass is calculated through the phenomenological\ncharacteristics of the vacuum related to spontaneous breaking of chiral\nsymmetry in QCD. The mass is determined by the mixed quark-gluon condensate\n$\\langle \\bar q G q \\rangle$ that emerges in an appropriate correlation\nfunction due to triangle anomaly. For three flavor QCD in chiral limit we find\nthe numerical value for the $U(1)_A$ boson mass to be $M_0=310\\pm 50~{\\rm\nMeV}$.\n", " Relativistic definition of the phase of the electromagnetic field, involving\ntwo Lorentz invariants, based on the Riemann-Silberstein vector is adopted to\nextend our previous study [I. Bialynicki-Birula, Z. Bialynicka-Birula and C.\nSliwa, Phys. Rev. A 61, 032110 (2000)] of the motion of vortex lines embedded\nin the solutions of wave equations from Schroedinger wave mechanics to Maxwell\ntheory. It is shown that time evolution of vortex lines has universal features;\nin Maxwell theory it is very similar to that in Schroedinger wave mechanics.\nConnection with some early work on geometrodynamics is established. Simple\nexamples of solutions of Maxwell equations with embedded vortex lines are\ngiven. Vortex lines in Laguerre-Gaussian beams are treated in some detail.\n" ]

[ " We study using the Bethe-Salpeter formalism the excitation energies of the\nzincbacteriochlorinbacteriochlorin dyad, a paradigmatic photosynthetic complex.\nIn great contrast with standard timedependent density functional theory\ncalculations with (semi)local kernels, charge transfer excitations are\ncorrectly located above the intramolecular Q-bands transitions found to be in\nexcellent agreement with experiment. Further, the asymptotic Coulomb behavior\ntowards the true quasiparticle gap for charge transfer excitations at long\ndistance is correctly reproduced, showing that the present scheme allows to\nstudy with the same accuracy intramolecular and charge transfer excitations at\nvarious spatial range and screening environment without any adjustable\nparameter.\n", " We investigate the consistency of the fundamental governing equations of\nmotion in continuum mechanics. In the first step, we examine the governing\nequations for a system of particles, which can be considered as the discrete\nanalog of the continuum. Based on Newton's third law of action and reaction,\nthere are two vectorial governing equations of motion for a system of\nparticles, the force and moment equations. As is well known, these equations\nprovide the governing equations of motion for infinitesimal elements of matter\nat each point, consisting of three force equations for translation, and three\nmoment equations for rotation. We also examine the character of other first and\nsecond moment equations, which result in non-physical governing equations\nviolating Newton's third law of action and reaction. Finally, we derive the\nconsistent governing equations of motion in continuum mechanics within the\nframework of couple stress theory. For completeness, the original couple stress\ntheory and its evolution toward consistent couple stress theory are presented\nin true tensorial forms.\n", " We determine the hadronic light-by-light scattering contribution to the\nanomalous magnetic moment of the muon using the framework of Dyson-Schwinger\nand Bethe-Salpeter equations of QCD. Our result for the pseudoscalar ($\\pi^0,\n\\eta, \\eta'$) meson exchange diagram is commensurate with previous\ncalculations. In our calculation of the quark loop contribution we improve\nprevious approaches by implementing constraints due to gauge invariance. As a\nconsequence, our value $a_\\mu^{\\textrm{LBL;quarkloop}} = (136 \\pm 59)\\times\n10^{-11}$ is significantly larger. Taken at face value, this then leads to a\nrevised estimate of the total $a_\\mu=116\\,591\\,891.0(105.0)\\times 10^{-11}$.\n", " The gravitational field and the source-free electromagnetic field can be\nunified preliminarily by the equations in the Riemannian geometry, both are\ncontractions of im and ik, respectively. So it will be equivalent to the Yang\ngravitational equations. From this we can obtain the Lorentz equations of\nmotion, the first system and second source-free system of Maxwell field\nequations. This unification can be included in the gauge theory, and the\nunified gauge group is GL(2,C), which is just the same as the gauge group of\nthe Riemanian manifold.\n", " Duality is one of the oldest known symmetries of Maxwell equations. In recent\nyears the significance of duality symmetry has been recognized in superstrings\nand high energy physics and there has been a renewed interest on the question\nof local duality rotation invariance. In the present paper we re-visit global\nduality symmetry in the Maxwell action and delineate the ambiguous role of\ngauge invariance and time locality. We have recently demonstrated that local\nduality invariance in a Lorentz covariant form can be carried out in the\nMaxwell equations. In this paper it is shown that in the four-pseudo vector\nLagrangian theory of Sudbery a local duality generalization can be naturally\nand unambiguously implemented and the Euler-Lagrange equations of motion are\nconsistent with the generalized Maxwell field equations. It is pointed out that\nthe extension of Noether theorem in full generality for a vector action is an\nimportant open problem in mathematical physics. Physical consequences of this\ntheory for polarized light and topological insulators are also discussed.\n", " Sturm theory for second order differential equations was generalized to\nsystems and higher order equations with positive leading coefficient by several\nauthors. Here we propose a Sturm oscillation theorem for indefinite systems of\neven order and with Dirichlet boundary conditions having strongly indefinite\nleading term.\n", " Motivated by the uniqueness problem for monostable semi-wavefronts, we\npropose a revised version of the Diekmann and Kaper theory of a nonlinear\nconvolution equation. Our version of the Diekmann-Kaper theory allows 1) to\nconsider new types of models which include nonlocal KPP type equations (with\neither symmetric or anisotropic dispersal), non-local lattice equations and\ndelayed reaction-diffusion equations; 2) to incorporate the critical case\n(which corresponds to the slowest wavefronts) into the consideration; 3) to\nweaken or to remove various restrictions on kernels and nonlinearities. The\nresults are compared with those of Schumacher (J. Reine Angew. Math. 316:\n54-70, 1980), Carr and Chmaj (Proc. Amer. Math. Soc. 132: 2433-2439, 2004), and\nother more recent studies.\n", " A modification of the Maxwell equations due to the presence of a\ngravitational field was formerly proposed for a scalar theory with a preferred\nreference frame. With this modification, the electric charge is not conserved.\nThe aim of the present work was to numerically assess the amount of charge\nproduction or destruction. We propose an asymptotic scheme for the\nelectromagnetic field in a weak and slowly varying gravitational field. This\nscheme is valid independently of the theory and the \"gravitationally-modified\"\nMaxwell equations. Then we apply this scheme to plane waves and to a group of\nHertzian dipoles in the scalar ether theory. The predicted amounts of charge\nproduction/destruction discard the formerly proposed gravitationally-modified\nMaxwell equations. The theoretical reason for that is the assumption that the\ntotal energy tensor is the sum of the energy tensor of the medium producing the\nelectromagnetic (e.m.) field and the e.m. energy tensor. This means that an\nadditional, \"interaction\" tensor has to be present. With this assumption, the\nstandard Maxwell equations in a curved spacetime, which predict charge\nconservation, are compatible with the investigated theory. We find that the\ninteraction energy might contribute to the dark matter.\n", " In a covariant quark-diquark model, we investigate the rare decay of\n$\\Lambda_b \\rightarrow n l^+ l^-$ and $\\Lambda_b \\rightarrow n \\gamma$ in the\nBethe-Salpeter equation approach. In this model the baryons are treated as\nbound states of a constituent quark and a diquark interacting via a gluon\nexchange and the linear confinement. We find that the ratio of form factors $R$\nis varies from $-0.90$ to $-0.25$ and the branching ratios $Br(\\Lambda_b\n\\rightarrow n l^+ l^-)\\times 10^{8}$ are $6.79(l=e),~4.08 ~ (l=\\mu),~2.9\n~(l=\\tau) $ and the branching ratio $Br(\\Lambda_b \\rightarrow \\gamma)\\times\n10^{7} )=3.69$ in central values of parameters.\n", " We study a noncommutative theory of gravity in the framework of torsional\nspacetime. This theory is based on a Lagrangian obtained by applying the\ntechnique of dimensional reduction of noncommutative gauge theory and that the\nyielded diffeomorphism invariant field theory can be made equivalent to a\nteleparallel formulation of gravity. Field equations are derived in the\nframework of teleparallel gravity through Weitzenbock geometry. We solve these\nfield equations by considering a mass that is distributed spherically\nsymmetrically in a stationary static spacetime in order to obtain a\nnoncommutative line element.This new line element interestingly reaffirms the\ncoherent state theory for a noncommutative Schwarzschild black hole. For the\nfirst time, we derive the Newtonian gravitational force equation in the\ncommutative relativity framework, and this result could provide the possibility\nto investigate examples in various topics in quantum and ordinary theories of\ngravity.\n", " A simple general relativity theory for objects moving in gravitational fields\nis developed based on studying the behavior of an atom in a gravitational\nfield. The theory is applied to calculate the satellite time dilation, light\ndeflection by the sun, and Mercury's precession with an error of less than 1%.\nThe approach to the new theory introduced here is radically different from the\ngeometric approach used by Einstein's general relativity. The theory is field\nbased where the potential energy of a system of masses can be easily calculated\nand the force can be found as the gradient of the potential field in analogy to\nthe Newtonian mechanics. The resulting field equations become the traditional\nNewton's equations when week gravitational effects are present. The theory\ncomplies with all the known experimental results such as the gravitational time\ndilation and faster light speeds higher in the gravitational field. The special\nrelativity theory of an object moving without experiencing gravitational fields\ncan be derived directly from the gravitational field equations introduced here.\nThe theory introduced here has crucial differences to Einstein's general\nrelativity theory. For example, the gravitational field cannot accelerate an\nobject to higher than the speed of light and the event horizon of a black hole\n(where light cannot escape) has to be of zero radius, essentially meaning that\nlight can escape any object unless the object has infinite density. Another\nprimary consequence of this study is that the principle of equivalence of\ngravitational and inertial mass has only limited validity and a new definition\nof gravitational mass is given here.\n", " In this paper, we study different types of weighted Besov and\nTriebel-Lizorkin spaces with variable smoothness. The function spaces can be\ndefined by means of the Littlewood-Paley theory in the field of Fourier\nanalysis, while there are other norms arising in the theory of partial\ndifferential equations such as Sobolev-Slobodeckij spaces. It is known that two\nnorms are equivalent when one considers constant regularity function spaces\nwithout weights. We show that the equivalence still holds for variable\nsmoothness and weights, which is accomplished by making use of shifted maximal\nfunctions, Peetre's maximal functions, and the reverse H\\\"older inequality.\nMoreover, we obtain a weighted regularity estimate for time-fractional\nevolution equations and a generalized Sobolev embedding theorem without\nweights.\n", " Hybrid organic/inorganic-semiconductor systems have important applications in\nboth molecular electronics and in photo-responsive materials. The\ncharacterization of the interface and of the electronic excited-states of these\nhybrid systems remains a challenge for state-of-the-art computational methods,\nas the systems of interest are large. In the present investigation, we present\nfor the first time a many-body Green's function Bethe-Salpeter investigation of\na series of photochromic molecules adsorbed onto TiO2 nanoclusters. Based on\nthese studies, the performance of TD-DFT is assessed. Using a state-of-the-art\ncomputational protocol, the photochromic properties of different hybrid systems\nare assessed. This work shows that qualitatively different conclusions can be\nreached with TD-DFT relying on various exchange-correlation functionals for\nsuch organic/inorganic interfaces, and paves the way to more accurate\nsimulation of many materials.\n", " The strong OZI allowed decays of the first radial excited states $B(2S)$ and\n$D(2S)$ are studied in the instantaneous Bethe-Salpeter method, and by using\nthese OZI allowed channels we estimate the full decay widths:\n$\\Gamma_{B^0(2S)}=24.4$ MeV, $\\Gamma_{B^+(2S)}=23.7$ MeV,\n $\\Gamma_{D^0(2S)}=11.3$ MeV and $\\Gamma_{D^+(2S)}=11.9$ MeV.\n We also predict the masses of them: $M_{B^0(2S)}=5.777$ GeV,\n$M_{B^+(2S)}=5.774$ GeV, $M_{D^0(2S)}=2.390$ GeV and $M_{D^+(2S)}=2.393$ GeV.\n", " The paper studies the inferences of wave equations for electromagnetic fields\nwhen there are gravitational fields at the same time. In the description with\nthe algebra of octonions, the inferences of wave equations are identical with\nthat in conventional electromagnetic theory with vector terminology. By means\nof the octonion exponential function, we can draw out that the electromagnetic\nwaves are transverse waves in a vacuum, and rephrase the law of reflection,\nSnell's law, Fresnel formula, and total internal reflection etc. The study\nclaims that the theoretical results of wave equations for electromagnetic\nstrength keep unchanged in the case for coexistence of gravitational and\nelectromagnetic fields. Meanwhile the electric and magnetic components of\nelectromagnetic waves can not be determined simultaneously in electromagnetic\nfields.\n", " The field equations of Kaluza-Klein (KK) theory have been applied in the\ndomain of cosmology. These equations are solved for a flat universe by taking\nthe gravitational and the cosmological constants as a function of time t. We\nuse Taylor's expansion of cosmological function, $\\Lambda(t)$, up to the first\norder of the time $t$. The cosmological parameters are calculated and some\ncosmological problems are discussed.\n", " Here we discuss the presence of CPT-even Lorentz violation (LV) in the\npresence of a deformed Heisenberg algebra that leads to a minimum length (ML).\nWe consider the case of a Maxwell lagrangian modified by the presence of a\n$K_{F}$ CPT-even LV theory and ML. We then derive a set of modified Maxwell\nequations in the cases of LV and ML and only ML. We verified that in the case\nof electromagnetic waves in the vacuum the presence of ML does not change the\nconsequences of LV. On the other hand, in a material media ML changes the whole\nset of equations that can lead to important effects with respect to the usual\nequations. We also considered the more general case including LV and the\nmodified equations in terms of matter fields. We then derived the refractive\nindex as a function of the matter fields depending on LV and ML, and in\nparticular we showed the behaviour of the refractive index with respect to the\nnon-commutative parameter.\n", " It is shown that the set of equations known as Maxwell's equations perfectly\ndescribe two very different systems: (1) the usual electromagnetic phenomena in\nvacuum or in the matter and (2) the deformation of isotropic solid lattices,\ncontaining topological defects as dislocations and disclinations, in the case\nof constant and homogenous expansion. The analogy between these two physical\nsystems is complete, as it is not restricted to one of the two Maxwell's\nequation couples in the vacuum, but generalized to the two equation couples as\nwell as to the diverse phenomena of dielectric polarization and magnetization\nof matter, just as to the electrical charges and the electrical currents. The\neulerian approach of the solid lattice developed here includes Maxwell's\nequations as a special case, since it stems from a tensor theory, which is\nreduced to a vector one by contraction on the tensor indices. Considering the\ntensor aspect of the eulerian solid lattice deformation theory, the analogy can\nbe extended to other physical phenomena than electromagnetism, a point which is\nshortly discussed at the end of the paper.\n", " The group of Conservative transformations is an enlargement of the group of\ndiffeomorphisms which leads to a richer geometry than that of general\nrelativity. The field variables of the theory are the usual orthonormal tetrads\nand also internal space tetrads. Using the fundamental geometric object which\nis the curvature vector, an appropriate Lagrangian has been defined for both\nfree-field and fields with sources. Solutions to the corresponding field\nequations have been developed. In this paper we use the static spherically\nsymmetric tetrad field with sources to model the universe. Our fundamental\nassumption is that the total density comprised of both ordinary and dark matter\nshould be constant. The resulting model with one adjustable parameter predicts\nthat ordinary matter is approximately 77\\% of the total mass content, but this\npercentage is near 0\\% for regions near the center of the universe. The space\nis approximately isotropic for $r$ near zero. The radial and tangential\npressures are negative and unequal. The redshift is also modeled without the\nexpanding universe and an explanation of the value of the cosmological constant\nis given. Equations governing particle motion are also derived which can\nproduce a repulsive effect and produce even larger redshifts. Finally, the\ncosmic microwave background and its anisotropies are addressed and heuristic\narguments are given that suggest that our theory is not inconsistent with these\nobservations. These results add further confirmation that the theory developed\nby Pandres is the fundamental theory of physics.\n", " We examine the equilibrium properties of hot, dilute, non-relativistic\nplasmas. The partition function and density correlation functions of a\nclassical plasma with several species are expressed in terms of a functional\nintegral over electrostatic potential distributions. The leading order,\nfield-theoretic tree approximation automatically includes the effects of Debye\nscreening. Subleading, one-loop corrections are easily evaluated. The two-loop\ncorrections, however, have ultraviolet divergences. These correspond to the\nshort-distance, logarithmic divergence which is encountered in the spatial\nintegral of the Boltzmann exponential when it is expanded to third order in the\nCoulomb potential. Such divergences do not appear in the underlying quantum\ntheory --- they are rendered finite by quantum fluctuations. We show how such\ndivergences may be removed and the correct finite theory obtained by\nintroducing additional local interactions in the manner of modern effective\nquantum field theories. We obtain explicit results for density-density\ncorrelation functions through two-loop order and thermodynamic quantities\nthrough three-loop order. The induced couplings are shown to obey\nrenormalization group equations, and these equations are used to characterize\nall leading logarithmic contributions in the theory. A linear combination of\npressure and energy and number densities is shown to be described by a\nfield-theoretic anomaly. The effective theory allows us to evaluate very easily\nthe algebraic long-distance decay of density correlation functions.\n", " On spacetimes that are not time orientable we construct a U(1) bundle to\nmeasure the twisting of the time axis. This single assumption, and simple\nconstruction, gives rise to Maxwell's equations of electromagnetism, the\nLorentz force law and the Einstein-Maxwell equations for electromagnetism\ncoupled to General relativity. The derivations follow the Kaluza Klein theory,\nbut with the constraints required for connections on a U(1) bundle rather than\nfive spacetime dimensions. The non time orientability is seen to justify and\nconstrain Kaluza Klein theories exactly as required to unify gravitation with\nelectromagnetism. Unlike any other schemes, apparent net electric charges arise\nnaturally because the direction of the electric field reverses along a time\nreversing path. The boundary of a time reversing region can therefore have a\nnet electric flux and appear exactly as a region containing an electric charge.\nThe treatment is purely classical, but motivated by links between acausal\nstructures and quantum theory.\n", " Torsion propagation and torsion-spin coupling are studied in the perspective\nof the Velo-Zwanziger method of analysis; specifically, we write the most\nextensive dynamics of the torsion tensor and the most exhaustive coupling that\nis permitted between torsion and spinors, and check the compatibility with\nconstraints and hyperbolicity and causality of field equations: we find that\nsome components of torsion and many terms of the torsion-spin interaction will\nbe restricted away and as a consequence we will present the most general theory\nthat is compatible with all restrictions.\n", " We present a Hilbert space perspective to homogenization of standard linear\nevolutionary boundary value problems in mathematical physics and provide a\nunified treatment for (non-)periodic homogenization problems in thermodynamics,\nelasticity, electro-magnetism and coupled systems thereof. The approach permits\nthe consideration of memory problems as well as differential-algebraic\nequations. We show that the limit equation is well-posed and causal. We rely on\ntechniques from functional analysis and operator theory only.\n", " Scalar composite boson masses have been computed in QCD and Technicolor\ntheories with the help of the homogeneous Bethe-Salpeter equation (BSE),\nresulting in a scalar mass that is twice the dynamically generated fermion or\ntechnifermion mass ($m_{dyn}$). We show that in the case of walking (or\nquasi-conformal) technicolor theories, where the $m_{dyn}$ behavior with the\nmomenta may be quite different from the one predicted by the standard operator\nproduct expansion, this result is incomplete and we must consider the effect of\nthe normalization condition of the BSE to determine the scalar masses. We\ncompute the composite Higgs boson mass for several groups with technifermions\nin the fundamental and higher dimensional representations and comment about the\nexperimental constraints on these theories, which indicate that models based on\nwalking theories with fermions in the fundamental representation may, within\nthe limitations of our approach, have masses quite near the actual direct\nexclusion limit.\n", " We investigate compatibility between the stochastic infrared (IR) resummation\nof light test fields on inflationary spacetimes and renormalisation group\nrunning of the ultra-violet (UV) physics. Using the Wilsonian approach, we\nderive improved stochastic Langevin and Fokker-Planck equations which\nconsistently include the renormalisation group effects. With the exception of\nstationary solutions, these differ from the naive approach of simply replacing\nthe classical potential in the standard stochastic equations with the\nrenormalisation group improved potential. Using this new formalism, we\nexemplify the IR dynamics with the Yukawa theory during inflation, illustrating\nthe differences between the consistent implementation of the UV running and the\nnaive approach.\n", " We present a simple proof of the continuity, in the sense distributions, of\nthe minors of the differential matrices of mappings belonging to grand Sobolev\nspaces. Such function spaces were introduced in connection with a problem on\nminimal integrability of the Jacobian and are useful in certain aspects of\ngeometric function theory and partial differential equations.\n", " In this paper we study the approximate controllability and existence of\noptimal control of impulsive fractional semilinear delay differential equations\nwith non-local conditions. We use Sadovskii's fixed point theorem, semigroup\ntheory of linear operators and direct method for minimizing a functional to\nestablish our results. At the end we give an example to illustrate our\nanalytical findings.\n", " In this paper we study the properties of vortexes, as systems specific to the\nAcoustic World, using both hydrodynamic theory and the corresponding\nhydrodynamic Maxwell equations. According to this study, it follows that the\nvortex behaves like an acoustic dipole that has intrinsic/internal angular\nmomentum. The system of two identical vortices also has orbital angular\nmomentum and behaves, at distances much greater than the distance between the\naxes of the vortices, as a single vortex. With the help of Maxwell's\nhydrodynamic equations for the vortex we deduced the force between two vortices\nand obtained the expression of the equivalent mass of the vortex and the\npermittivity of the electroacoustic field. We also obtained and interpreted the\nexpression for the energy density of the acoustic field. The density and\npressure variations induced by the vortex cause the change in the propagation\nspeed of the acoustic waves and the acoustic lensing property of the vortex.\n", " The notion of fractional dynamics is related to equations of motion with one\nor a few terms with derivatives of a fractional order. This type of equation\nappears in the description of chaotic dynamics, wave propagation in fractal\nmedia, and field theory. For the fractional linear oscillator the physical\nmeaning of the derivative of order $\\alpha<2$ is dissipation. In systems with\nmany spacially coupled elements (oscillators) the fractional derivative, along\nthe space coordinate, corresponds to a long range interaction. We discuss a\nmethod of constructing a solution using an expansion in $\\epsilon=n-\\alpha$\nwith small $\\epsilon$ and positive integer $n$. The method is applied to the\nfractional linear and nonlinear oscillators and to fractional Ginzburg-Landau\nor parabolic equations.\n", " We present the public version of hi_class (www.hiclass-code.net), an\nextension of the Boltzmann code CLASS to a broad ensemble of modifications to\ngeneral relativity. In particular, hi_class can calculate predictions for\nmodels based on Horndeski's theory, which is the most general scalar-tensor\ntheory described by second-order equations of motion and encompasses any\nperfect-fluid dark energy, quintessence, Brans-Dicke, $f(R)$ and covariant\nGalileon models. hi_class has been thoroughly tested and can be readily used to\nunderstand the impact of alternative theories of gravity on linear structure\nformation as well as for cosmological parameter extraction.\n", " The Standard Model is an up-to-date theory that best summarizes current\nknowledge in particle physics. Although some problems still remain open, it\nrepresents the leading model which all physicists refer to. One of the pillars\nwhich underpin the Standard Model is represented by the Lorentz invariance of\nthe equations that form its backbone. These equations made it possible to\npredict the existence of particles and phenomena that experimental physics had\nnot yet been able to detect. The first hint of formulating a fundamental theory\nof particles can be found in the 1932 Majorana equation, formulated when\nelectrons and protons were the only known particles. Today we know that parts\nof the hypotheses set by Majorana were not correct, but his equation hid\nconcepts that are found in the Standard Model. In this study, the Majorana\nequation is revisited and solved for free particles. The time-like, light-like\nand space-like solutions, represented by infinite-component wave functions, are\ndiscussed. Furthermore, by introducing subsidiary conditions on the mass term,\nit is possible to quantize both the fermionic and the bosonic towers, obtaining\nthe mass spectrum of the entire family of charged leptons, baryons and mesons.\n", " The short history of L.Kantorovich's transport problem and his metric in the\nframework of his activity in mathmatics and economics during a long and\ndifficult period. We did not mention the recent impetuious developement of\napplication of the transport probelm to differential equations etc., and\nconcentrated on three applications not known to wide audience - in measure and\nergodic theory, and metric geometry and classification of mertric spaces with\nmeasures.\n This is a detale version of the talk of author on the Conference devoted to\n 90-th annyversary of Leonid.V.Kantorovich which was at St.Petersbutrg Euler\nInternational Mathematical Institute in January 2004.\n", " This paper intends to lay the theoretical foundation for the method of\nfunctional maps, first presented in 2012 by Ovsjanikov, Ben-Chen, Solomon,\nButscher and Guibas in the field of the theory and numerics of maps between\nshapes. We show how to analyze this method by looking at it as an application\nof the theories of composition operators, of matrix representa- tion of\noperators on separable Hilbert spaces, and of the theory of the Finite Section\nMethod. These are three well known fruitful topics in functional analysis. When\napplied to the task of modelling of correspondences of shapes in\nthree-dimensional space, these concepts lead directly to functional maps and\nits associated functional matrices. Mathematically spoken, functional maps are\ncomposition operators between two-dimensional manifolds, and functional\nmatrices are infinite matrix representations of such maps. We present an\nintroduction into the notion and theoretical foundation of the functional\nanalytic framework of the theory of matrix repre- sentation, especially of\ncomposition operators. We will also discuss two numerical methods for solving\nequations with such operators, namely, two variants of the Rectangular Finite\nSection Method. While one of these, which is well known, leads to an\noverdetermined system of linear equations, in the second one the minimum-norm\nsolution of an underdetermined system has to be computed. We will present the\nmain convergence results related to these methods.\n", " General Theory of Relativity and Quantum theory gives two different\ndescription of the same mother nature in the big and small scale respectively.\nMathematical languages of these two theories are entirely different, one is\ngeometric while the other one is probabilistic. This curious feature makes the\nmerging of these theories (quantum-gravity) considerably difficult. In this\nmanuscript, we explore quantum-gravity in a dynamical system perspective. For\nexample, in the standard quantum theory, the wave equation is a linear partial\ndifferential equation while in General Theory of relativity, the field\nequations are highly nonlinear. A classical dynamical system can show very rich\nphase-space structures which is absent in a linear partial differential\nequation. In order to incorporate gravitational corrections one can think about\nthe nonlinear extensions of quantum theory. In this manuscript such an attempt\nto quantum gravity is explored. Dissipative and forcing corrections are found\nin the newly formulated quantum equation and it's physical interpretations are\ngiven.\n", " The main purpose of this paper is to treat semigroups properties, like norm\ncontinuity, compactness and differentiability for perturbed semigroups in\nBanach spaces. In particular, we investigate three large classes of\nperturbations, Miyadera-Voigt, Desch-Schappacher and Staffans-Weiss\nperturbations. Our approach is mainly based on feedback theory of Salamon-Weiss\nsystems. Our results are applied to abstract boundary integro-differential\nequations in Banach spaces.\n", " HELAC-PHEGAS is a FORTRAN based package that is able to compute automatically\nand efficiently tree-order helicity amplitudes and cross sections for arbitrary\nscattering processes within the standard electroweak theory and QCD. The\nalgorithm for the amplitude computation, HELAC, exploits the virtues of the\nDyson-Schwinger equations. The phase-space generation algorithm, PHEGAS,\nconstructs all possible kinematical mappings dictated by the amplitude under\nconsideration. Combined with mutichannel self-optimized Monte Carlo integration\nit results to efficient cross section evaluation.\n", " In 1945 Einstein concluded that [1]: 'The present theory of relativity is\nbased on a division of physical reality into a metric field (gravitation) on\nthe one hand, and into an electromagnetic field and matter on the other hand.\nIn reality space will probably be of a uniform character and the present theory\nbe valid only as a limiting case. For large densities of field and of matter,\nthe field equations and even the field variables which enter into them will\nhave no real significance.'. The dichotomy can be resolved by introducing a\nscalar field/potential algebraically related to the Ricci tensor for which the\ncorresponding metric is free of additional singularities. Hence, although a\nfundamentally nonlinear theory, the scalar field/potential provides an analytic\nframework for interacting particles; described by linear superposition. The\nstress tensor for the scalar field includes both the sources of and the\nenergy-momentum for the gravitational field, and has zero covariant and\nordinary divergence. Hence, the energy-momentum for the gravitational field and\nsources are conserved. The theory's predictions agree with the experimental\nresults for General Relativity. By introducing the corresponding Lagrangian to\nanalytic mechanics, what is experimentally known for GR can be accounted for.\n", " The turbulence field is stacked on the laminar flow. In this research, the\nlaminar flow is described as a macro deformation which forms an instant\ncurvature space. On such a curvature space, the turbulence is viewed as a micro\ndeformation. So, the fluid flow is described by the geometrical field theory of\nfinite deformation. Based on the Lagrange mechanics and the deformation energy\nconcept, using the Least Action Principle, the Euler-Lagrange motion equations\nare obtained. According to A E Green formulation, the stress concept is\nintroduced by deformation tensor. The fluid motion is described by the\nmultiplication of a macro deformation tensor and a micro deformation tensor. By\nthis way, the geometrical field of fluid motion is well constructed. Then, the\nspatial derivative of deformation energy is expressed by the gradient of\ndeformation tensors. By this way, the deformation energy related items in the\nEuler-Lagrange motion equations are expressed by the stress tensor and\ndeformation tensor. The obtained Euler-Lagrange motion equations, then, are\ndecomposed into average deformation equations and turbulence equations. For\nseveral special cases, the new results are compared with the conventional\nNavier-Stokes equation with Reynolds stress modification.The comparisons also\nshow that the Bernoulli Equation is a natural precondition for the conventional\nNavier-Stokes equation.Generally, the turbulence wave is an inward-traveling\nwave. Unlike the normal outward-traveling wave related with the average\ndeformation, the inward-traveling wave is the intrinsic feature of turbulence.\nSo, the turbulence is well defined by the equations obtained in this research.\nFor several typical cases, the simplified turbulence wave equations are given\nout with simple discussion.\n", " We revisit homogenisation theory for Friedrichs systems. In particular, we\nshow that $G$-compactness can be obtained under severely weaker assumptions\nthan in the original work of Burazin and Vrdoljak (2014). In this way we extend\nthe applicability of $G$-compactness results for Friedrichs systems to\nequations that yield memory effects in the homogenised limit and detour any\nusage of compactness techniques previously employed.\n", " There are unsolved problems in physics leading to difficulties with Maxwell\nequations that are not removed by and not directly associated with quantum\nmechanics. Recently a number of extended and modified theories have therefore\nbeen put forward. The two theories by the authors (AK and BL) are related to\nthis state. The first theory consists of a revised formulation of\nelectromagnetics with an electric vacuum charge density (BL), and the second is\na unified theory of Bivacuum, matter and fields (AK). In the first theory by BL\na space-charge current density arises from the condition of Lorentz invariance,\nin addition to the conventional displacement current. The electron charge is\ndeduced from the theory, and it deviates by only one percent from its\nexperimental value. Two individual photon models are further investigated.\n In the second theory by AK a Bivacuum model is a consequence of development\nof Dirac theory, pointing to equal probability of positive and negative energy.\nThe dynamic mechanism of [corpuscle (C) - wave (W)] duality is proposed. It\nunifies the extended relativity and QM. For details see:\nhttp://arxiv.org/abs/physics/0003001. The theories of BL and AK reinforce each\nother, by having several points in common. There are the similarities with the\nDirac theory, symmetry with respect to the concepts of matter and antimatter,\nthe non-appearance of magnetic monopoles, a space- charge vacuum current in\naddition to the displacement current, models of the leptons having an internal\nvortex-like structure and force balance, photon models with a small but nonzero\nrest mass and a slightly reduced velocity of propagation, and particle-wave\nproperties, explaining self-interference even of single electrons or photons.\n", " We find and discuss asymptotic formulas for orthonormal polynomials\n$P_{n}(z)$ with recurrence coefficients $a_{n}, b_{n}$. Our main goal is to\nconsider the case where off-diagonal elements $a_{n}\\to\\infty$ as $n\\to\\infty$.\nFormulas obtained are essentially different for relatively small and large\ndiagonal elements $b_{n}$.\n Our analysis is intimately linked with spectral theory of Jacobi operators\n$J$ with coefficients $a_{n}, b_{n}$ and a study of the corresponding second\norder difference equations.\n We introduce the Jost solutions $f_{n}(z)$, $n\\geq -1$, of such equations by\na condition for $n\\to\\infty$ and suggest an Ansatz for them playing the role of\nthe semiclassical Liouville-Green Ansatz for solutions of the Schr\\\"odinger\nequation.\n This allows us to study the spectral structure of Jacobi operators and their\neigenfunctions $P_{n}(z)$ by traditional methods of spectral theory developed\nfor differential equations. In particular, we express all coefficients in\nasymptotic formulas for $P_{n}(z)$ as $n \\to\\infty$ in terms of the Wronskian\nof the solutions $ P_{n} (z) $ and $ f_{n} (z)$. The formulas obtained for\n$P_{n}(z)$ generalize the asymptotic formulas for the classical Hermite\npolynomials where $a_{n}=\\sqrt{(n+1)/2}$ and $b_{n}=0$.\n", " How the energy conservation law is obeyed by the universe during its\nevolution is an important but not yet unanimously resolved question. Does the\nuniverse have a constant mass during its evolution or has its mass been\nincreasing with its radius? Here, we evaluate the two contending propositions\nwithin the context of the Friedmann equations and the standard big bang theory.\nWe find that though both propositions appeal to the Friedmann equations for\nvalidity, an increasing mass with increasing radius is more in harmony with the\nthermal history of the big bang model. In addition, temperature and flatness\nproblems that plague the constant mass proposal are mitigated by the increasing\nmass with radius proposal. We conclude that the universe has been increasing in\nmass and radius in obedience to the energy conservation law.\n", " Historically, the thermodynamic behavior of gasses was described first and\nthe derived equations were adapted to solids. It is suggested that the current\nthermodynamic description of solid phase is still incomplete because the\nisothermal work done on or by the system is not counted in the internal energy.\nIt is also suggested that the isobaric work should not be deducted from the\ninternal energy because the system does not do work when it expands. Further\nmore it is suggested that Joule postulate regarding the mechanical equivalency\nof heat -the first law of thermodynamics- is not universal and not applicable\nto elastic solids. The equations for the proposed thermodynamic description of\nsolids are derived and tested by calculating the internal energies of the\nsystem using the equation of state of MgO. The agreement with theory is good.\n", " We construct a manifestly Machian theory of gravitation on the foundation\nthat information in the universe cannot be destroyed (Landauer's principle). If\nno bit of information in the Universe is lost, than the sum of the entropies of\nthe geometric and the matter fields should be conserved. We propose a local\ninvariant expression for the entropy of the geometric field and formulate a\nvariational principle on the entropic functional which produces entropic field\nequations. This information-theoretic approach implies that the geometric field\ndoes not exist in an empty of matter Universe, the material entropy is geometry\ndependent, matter can exchange information (entropy) with the geometric field\nand a quantum condensate can channel energy into the geometric field at a\nparticular coherent state. The entropic field equations feature a non-intuitive\ndirect coupling between the material fields and the geometric field, which acts\nas an entropy reservoir. Cosmological consequences such as the emergence of the\ncosmological constant as well as experimental consequences involving\ngravity-quantum condensate interaction are discussed. The energetic aspect of\nthe theory restores the repertoire of the classical General Relativity up to a\ndifferent coupling constant between the fields.\n", " We determine the leading Fock state light front wave functions (LFWFs) of the\npion and kaon via light front projections of the covariant Bethe-Salpeter wave\nfunction. Using these LFWFs we study the multi-dimensional images of the\nvalence quarks in the pion and kaon that are provided by their generalized\nparton distribution functions (GPDs) and transverse momentum dependent parton\ndistribution functions (TMDs). Moments of the GPDs are taken to obtain the\nelectromagnetic and gravitational form factors of the pion and kaon, and\ncomparisons to available experimental and lattice data are made. Highlights\nfrom this study include predictions that the mean-squared impact parameter for\nthe quarks in the pion and kaon are: $\\langle\n\\vec{b}_T^2\\rangle_{u}^\\pi=0.11$fm$^2$, $\\langle\n\\vec{b}_T^2\\rangle_{s}^K=0.08$fm$^2$, and $\\langle\n\\vec{b}_T^2\\rangle_{u}^K=0.13$fm$^2$, and therefore the $s$ quark in the kaon\nis much closer to the center of transverse momentum than the $u$ quark. From\nthe electromagnetic and gravitational form factors we find that the light-cone\nenergy radii are about 60\\% smaller than the light-cone charge radii for each\nquark sector in the pion and kaon. A quantitative measure of the importance of\nthe leading Fock state is obtained via comparison with a full DSE calculation\n(containing an infinite tower of Fock states) for the pion form factor.\n", " A universal quantum wave equation that yields Dirac, Klein-Gordon,\nSchrodinger and quantum heat equations is derived. These equations are related\nby complex transformation of space, time and mass. The new symmetry exhibited\nby these equations is investigated. The universal quantum equation yields Dirac\nequation in two ways: firstly by replacing the particle my $m_0$ by $im_0$, and\nsecondly by changing space and time coordinates by $it$ and $i\\vec{r}$,\nrespectively.\n", " We exhibit an alternative method for solving inhomogeneous second--order\nlinear ordinary dynamic equations on time scales, based on reduction of order\nrather than variation of parameters. Our form extends recent (and\nlong-standing) analysis on $\\R$ to a new form for difference equations, quantum\nequations, and arbitrary dynamic equations on time scales.\n", " A Lagrangian depending on geometric variables (metric, affine connection,\ngauge group generators) is given which maintains compatibility with General\nRelativity. It generates the dynamics for Electromagnetism and other Gauge\nFields along with Gravitation, at the time it gives a geometric foundation for\nthe stress-energy tensor of continuous matter. The geometric-invariance\nprinciple under this integration is exposed and the resulting field equations\nare obtained. The theory is developed over the tangent space of a\nfour-dimensional real manifold and the generators become those from the\nHomogenous Lorentz group.\n", " The new great development in Physics could be related to the excited progress\nof a new mathematics: ternary theory of numbers, ternary Pithagor theorem and\nternary complex analysis, ternary algebras and symmetries, ternary Clifford\nalgebras,ternary differential geometry, theory of the differential wave\nequations of the higher degree n>2 and etc. Especially, we expect the powerful\ninfluence of this progress into the Standard Model (SM) and beyond, into high\nenergy neutrino physics, Gravity and Cosmology. This can give the further\ndevelopment in the understanding of the Lorentz symmetry and matter-antimatter\nsymmetry, the geometrical origin of the gauge symmetries of the Standard Model,\nof the 3-quark-lepton family and neutrino problems, dark matter and dark energy\nproblems in Cosmology.\n The new ambient geometry can be related to a new space-time symmetry leading\nat high energies to generalization of the Special Theory of Relativity. We\nrelated the future of this development with C_n numbers, n-algebras (n>2) and\ncorresponding geometrical objects. We will discuss the following results - C_n\ncomplexification of R^n spaces - C_n structure and the invariant surfaces - C_n\nholomorphicity and harmonicity - The link between C_n holomorphicity and the\norigin of spin 1/n - New geometry and N-ary algebras/symmetries - Root system\nof a new ternary TU(3) algebra - N-ary Clifford algebras - Ternary 9-plet and\n27-plet number surfaces\n", " Admissible point transformations of classes of $r$th order linear ordinary\ndifferential equations (in particular, the whole class of such equations and\nits subclasses of equations in the rational form, the Laguerre-Forsyth form,\nthe first and second Arnold forms) are exhaustively described. Using these\nresults, the group classification of such equations is revisited within the\nalgebraic approach in three different ways.\n" ]

[ " We present a generalization of the Fefferman-de la Llave decomposition of the\nCoulomb potential to quite arbitrary radial functions $V$ on $R^n$ going to\nzero at infinity. This generalized decomposition can be used to extend previous\nresults on $N$-body quantum systems with Coulomb interaction to a more general\nclass of interactions. As an example of such an application we derive the high\ndensity asymptotics of the ground state energy of jellium with Yukawa\ninteraction in the thermodynamic limit, using a correlation estimate by Graf\nand Solovej.\n", " We argue that in the context of string theory a large number N of connected\ndegenerate vacua that mix will lead to a ground state with much lower energy,\nessentially because of the standard level repulsion of quantum theory for the\nwavefunction of the Universe. We imagine a history where initial quantum\nfluctuations give an energy density $\\sim m_{susy}^2m_{Pl}^2$, but the universe\nquickly cascades to an energy density $\\sim m_{susy}^2m_{Pl}^2/N$. Then at\nvarious phase transitions there are large contributions to the energy density\nand rearrangement of levels, followed again by a rapid cascade to the ground\nstate or near it. If this mechanism is correct, the ground state of the theory\ndescribing our world would be a superposition of a large number of connected\nstring vacua,with shared superselection sets of properties such as three\nfamilies etc. The observed value of the cosmological constant in terms of the\nPlanck mass, the scale of supersymmetry breaking and the number of connected\nstring vacua.\n", " We investigate on the procedure of extracting a \"spectral density\" from mixed\nQM/MM calculations and employing it in open quantum systems models. In\nparticular, we study the connection between the energy gap correlation function\nextracted from ground state QM/MM and the bath spectral density used as input\nin open quantum system approaches. We introduce a simple model which can give\nintuition on when the ground state QM/MM propagation will give the correct\nenergy gap. We also discuss the role of higher order correlators of the\nenergy-gap fluctuations which can provide useful information on the bath.\nFurther, various semiclassical corrections to the spectral density, are applied\nand investigated. Finally, we apply our considerations to the photosynthetic\nFenna-Matthews-Olson complex. For this system, our results suggest the use of\nthe Harmonic prefactor for the spectral density rather than the Standard one,\nwhich was employed in the simulations of the system carried out to date.\n", " In this letter we report the singlet ground state structure of the full\ncarotenoid peridinin by means of variational Monte Carlo (VMC) calculations.\nThe VMC relaxed geometry has an average bond length alternation of 0.1165(10)\n{\\AA}, larger than the values obtained by DFT (PBE, B3LYP and CAM-B3LYP) and\nshorter than that calculated at the Hartree-Fock (HF) level. TDDFT and EOM-CCSD\ncalculations on a reduced peridinin model confirm the HOMO-LUMO major\ncontribution of the Bu+-like (S2) bright excited state. Many Body Green's\nFunction Theory (MBGFT) calculations of the vertical excitation energy of the\nBu+-like state for the VMC structure (VMC/MBGFT) provide excitation energy of\n2.62 eV, in agreement with experimental results in n-hexane (2.72 eV). The\ndependence of the excitation energy on the bond length alternation in the MBGFT\nand TDDFT calculations with different functionals is discussed.\n", " Hanging about a hypothetical connections between the ground state vector for\nsome special spin systems and the alternating-sign matrices, we have found a\nnumerical evidence for the fact that the numbers of the states of the fully\npacked loop model with fixed link-patterns coincide with the components of the\nground state vector of the dense O$(1)$ loop model considered by Batchelor, de\nGier and Nienhuis. Our conjecture generalizes in a sense the conjecture of\nBosley and Fidkowski, refined by Cohn and Propp, and proved by Wieland.\n", " We transpose work by K.Yajima and by T.Mizumachi to prove dispersive and\nsmoothing estimates for dispersive solutions of the linearization at a ground\nstate of a Nonlinear Schr\\\"odinger equation (NLS) in 2D. As an application we\nextend to dimension 2D a result on asymptotic stability of ground states of NLS\nproved in the literature for all dimensions different from 2.\n", " The LiFr diatomic represents a promising candidate for indirect laser cooling\nthat has not yet been investigated not theoretically or experimentally. The\npotential energy curves of the ground and low_lying excited states of the LiFr\nheteronuclear alkali metal dimer are calculated using the Fock_space\nrelativistic coupled cluster theory for the first time. A number of properties\nsuch as the electronic term energies, equilibrium internuclear distances,\ntransition and permanent dipole moments, sequences of vibrational energies,\nharmonic vibrational frequencies, Franck_Condon factors, and radiative\nlifetimes (including bound and free transitions) are predicted. The\nprobabilities of the two_step schemes (optical cycles) for the transfer process\nof the LiFr molecules from high excited vibrational states to the ground\nvibronic state are also predicted. The data obtained would be useful for laser\ncooling and spectral experiments with LiFr molecules.\n", " Application of quantum principles to living cells requires a new\napproximation of the full quantum mechanical description of intracellular\ndynamics. We discuss what principal elements any such good approximation should\ncontain. As one such element, the notion of \"Catalytic force\" Cf is introduced.\nCf is the effect of the molecular target of catalysis on the catalytic\nmicroenvironment that adjusts the microenvironment towards a state that\nfacilitates the catalytic act. This phenomenon is experimentally testable and\nhas an intriguing implication for biological organization and evolution, as it\namounts to \"optimization without natural selection of replicators\". Unlike the\nstatistical-mechanical approaches to self-organization, the Cf principle does\nnot encounter the problem of \"tradeoff between stability and complexity\" at the\nlevel of individual cell. Physically, the Cf is considered as a harmonic-like\nforce of reaction, which keeps the state of the cell close to the ground state,\ndefined here as a state where enzymatic acts work most efficiently. Ground\nstate is subject to unitary evolution, and serves as a starting point in a\ngeneral strategy of quantum description of intracellular processes, termed here\n\"Euclidean approach\". The next step of this strategy is transition from the\ndescription of ground state to that one of growing state, and we suggest how it\ncan be accomplished using arguments from the fluctuation-dissipation theorem.\nFinally, given that the most reliable and informative observable of an\nindividual cell is the sequence of its genome, we propose that the\nnon-classical correlations between individual molecular events at the single\ncell level could be easiest to detect using high throughput DNA sequencing.\n", " Nonadiabatic corrections in molecules composed of a few atoms are considered.\nIt is demonstrated that a systematic perturbative expansion around the\nadiabatic solution is possible, with the expansion parameter being the\nelectron-nucleus mass ratio to the power 3/4. Closed form formulae for the\nleading corrections to the wave function and to the energy are derived. Their\napplicability is demonstrated by a comparison of numerical results for the\nhydrogen molecule with the former nonadiabatic calculations and the\nexperimental values. Good agreement with the recent experiment is achieved for\nthe ground state dissociation energy of both H$_2$ and D$_2$.\n", " I briefly describe two motivations, two mechanisms and two possible tests of\nthe hypothesis that the physical parameters of the ground state of a theory can\nvary in different regions of the universe.\n", " We explore the connection between chaos, thermalization and ergodicity in a\nlinear chain of $N$ interacting dipoles. Starting from the ground state, and\nconsidering chains of different numbers of dipoles, we introduce single site\nexcitations with energy $\\Delta K$. The time evolution of the chaoticy of the\nsystem and the energy localization along the chain is analyzed by computing, up\nto very long times, the statistical average of the finite time Lyapunov\nexponent $\\lambda(t)$ and of the participation ratio $\\Pi(t)$. For small\n$\\Delta K$, the evolution of $\\lambda(t)$ and $\\Pi(t)$ indicates that the\nsystem becomes chaotic at roughly the same time as $\\Pi(t)$ reaches a steady\nstate. For the largest values of $\\Delta K$, the system becomes chaotic at an\nextremely early stage in comparison with the energy relaxation times. We find\nthat this fact is due to the presence of chaotic breathers that keep the system\nfar from equipartition and ergodicity. Finally, we show that the asymptotic\nvalues attained by the participation ratio $\\Pi(t)$ fairly corresponds to\nthermal equilibrium.\n", " We develop an efficient algorithm, coordinate descent FCI (CDFCI), for the\nelectronic structure ground state calculation in the configuration interaction\nframework. CDFCI solves an unconstrained non-convex optimization problem, which\nis a reformulation of the full configuration interaction eigenvalue problem,\nvia an adaptive coordinate descent method with a deterministic compression\nstrategy. CDFCI captures and updates appreciative determinants with different\nfrequencies proportional to their importance. We show that CDFCI produces\naccurate variational energy for both static and dynamic correlation by\nbenchmarking the binding curve of nitrogen dimer in the cc-pVDZ basis with\n$10^{-3}$ mHa accuracy. We also demonstrate the efficiency and accuracy of\nCDFCI for strongly correlated chromium dimer in the Ahlrichs VDZ basis and\nproduces state-of-the-art variational energy.\n", " It is possible to express all the strong and electromagnetic interactions of\nground state hadrons in terms of a single coupling constant and the constituent\nquark masses, using spin-flavour relativistic supermultiplet theory. Results\nare generally accurate to within 10%.\n", " We propose an efficient deterministic method to calculate the full\nconfiguration interaction (FCI) ground state energy. This method leverages the\nsparseness of the Lanczos basis vectors that span the Krylov subspace\nassociated with the Hamiltonian to rapidly calculate the lowest eigenvalue of\nthe effective Hamiltonian formed in this basis. By iteratively performing the\nspanning and diagonalization steps, this method is capable of rapidly reaching\nchemical accuracy for a variety of strongly correlated molecules, including\nstretched N$_2$ and C$_2$, in merely several tens of iterations. To accelerate\nour algorithm, we implement parallelized sparse matrix-sparse vector\nmultiplication, which features a novel hashing method that establishes a\none-to-one correspondence between the set of full configuration Slater\ndeterminants and a set of consecutive integers. A near-linear acceleration was\nobserved.\n", " Compton scattering is a key process shaping spectra formation and accretion\nflow dynamics in accreting strongly magnetized neutron stars. A strong magnetic\nfield affects the scattering cross section and makes it dependent on photon\nenergy, momentum, and polarization state. Using Monte Carlo simulations, we\ninvestigate statistical features of Compton scattering of polarized X-ray\nradiation in a strong magnetic field. Our analysis is focused on photon gas\nbehaviour well inside the scattering region. We take into account the resonant\nscattering at the fundamental cyclotron frequency, thermal distribution of\nelectrons at the ground Landau level, and bulk velocity of the electron gas. We\nshow that (i) the photons scattered around the cyclotron energy by the electron\ngas at rest tend to acquire the final energy close to the cyclotron one with a\nvery small dispersion measure; (ii) the redistribution of photons within the\nDoppler core of cyclotron resonance differs significantly from the complete\nredistribution; (iii) the efficiency of momentum transfer from photons to the\nelectron gas is affected by the temperature of electron gas both for photons at\ncyclotron energy and below it; (iv) the momentum transfer from photons to the\nelectron gas of non-zero bulk velocity is more efficient in the case of\nmagnetic scattering.\n", " Optomechanical systems close to their quantum ground state and nonlinear\nnanoelectromechanical systems are two hot topics of current physics research.\nAs high-reflectivity and low mass are crucial features to improve\noptomechanical coupling towards the ground state, we have designed, fabricated\nand characterized photonic crystal nanomembranes, at the crossroad of both\ntopics. Here we demonstrate a number of nonlinear effects with these membranes.\nWe first characterize the nonlinear behavior of a single mechanical mode and we\ndemonstrate its nonlocal character by monitoring the subsequent\nactuation-related frequency shift of a different mode. We then proceed to study\nthe underlying nonlinear dynamics, both by monitoring the phase-space\ntrajectory of the free resonator and by characterizing the mechanical response\nin presence of a strong pump excitation. We observe in particular the frequency\nevolution during a ring-down oscillation decay, and the emergence of a phase\nconjugate mechanical response to a weaker probe actuation. Our results are\ncrucial to understand the full nonlinear features of the PhC membranes, and\npossibly to look for nonlinear signatures of the quantum dynamics.\n", " Rovibrational quantum states in the $X^1\\Sigma_g^+$ electronic ground state\nof H$_2$ are prepared in the $v=13$ vibrational level up to its highest bound\nrotational level $J=7$, and in the highest bound vibrational level $v=14$ (for\n$J=1$) by two-photon photolysis of H$_2$S. These states are laser-excited in a\nsubsequent two-photon scheme into $F^1\\Sigma_g^+$ outer well states, where the\nassignment of the highest ($v,J$) states is derived from a comparison of\nexperimentally known levels in \\F, combined with \\emph{ab initio} calculations\nof \\X\\ levels. The assignments are further verified by excitation of\n$F^1\\Sigma_g^+$ population into autoionizing continuum resonances which are\ncompared with multi-channel quantum defect calculations. Precision\nspectroscopic measurements of the $F-X$ intervals form a test for the \\emph{ab\ninitio} calculations of ground state levels at high vibrational quantum numbers\nand large internuclear separations, for which agreement is found.\n", " We demonstrate high accuracy calculations for the HeH$^+$ molecule using\nnewly developed analytic formulas for two-center two-electron integrals with\nexponential functions. The Born-Oppenheimer potential for the ground electronic\n$^1\\Sigma^+$ state is obtained in the range of 0.1 -- 60 au with precision of\nabout $10^{-12}$ au. As an example at the equilibrium distance $r=1.463\\,283$\nau the Born-Oppenheimer potential amounts to $-2.978\\,708\\,310\\,771(1)$.\nObtained results lay the ground for theoretical predictions in HeH$^+$ with\nspectroscopic precision.\n", " We consider the Kirchhoff-type $p$-Laplacian Dirichlet problem containing the\nleft and right fractional derivative operators. By using the Nehari method in\ncritical point theory, we obtain the existence theorem of ground state\nsolutions for such Dirichlet problem.\n", " Here we examine O(n) systems with arbitrary two spin interactions (of\nunspecified range) within a general framework. We shall focus on\ntranslationally invariant interactions. In the this case, we determine the\nground states of the $O(n \\ge 2)$ systems. We further illustrate how one may\nestablish Peierls bounds for many Ising systems with long range interactions.\nWe study the effect of thermal fluctuations on the ground states and derive the\ncorresponding fluctuation integrals. The study of the thermal fluctuation\nspectra will lead us to discover a very interesting odd-even $n$\n(coupling-decoupling) effect. We will prove a generalized Mermin-Wagner-Coleman\n(integral divergence) theorem for all translationally invariant interactions in\ntwo dimensions with an analytic kernel in momentum space. We will show that\nmany three dimensional systems have smectic like thermodynamics. We will\nexamine the topology of the ground state manifolds for both translationally\ninvariant and spin glass systems. We conclude with a discussion of O(n) spin\ndynamics in the general case.\n", " Atoms made of a particle and an antiparticle are unstable, usually surviving\nless than a microsecond. Antihydrogen, made entirely of antiparticles, is\nbelieved to be stable, and it is this longevity that holds the promise of\nprecision studies of matter-antimatter symmetry. We have recently demonstrated\ntrapping of antihydrogen atoms by releasing them after a confinement time of\n172 ms. A critical question for future studies is: how long can anti-atoms be\ntrapped? Here we report the observation of anti-atom confinement for 1000 s,\nextending our earlier results by nearly four orders of magnitude. Our\ncalculations indicate that most of the trapped anti-atoms reach the ground\nstate. Further, we report the first measurement of the energy distribution of\ntrapped antihydrogen which, coupled with detailed comparisons with simulations,\nprovides a key tool for the systematic investigation of trapping dynamics.\nThese advances open up a range of experimental possibilities, including\nprecision studies of CPT symmetry and cooling to temperatures where\ngravitational effects could become apparent.\n", " We present results for masses of excited baryons from a quenched calculation\nwith Chirally Improved quarks at pion masses down to 350 MeV. Our analysis of\nthe correlators is based on the variational method. In order to provide a large\nbasis set for spanning the physical states, we use interpolators with different\nDirac structures and Jacobi smeared quark sources of different width. Our\nspectroscopy results for a wide range of ground state and excited baryons are\ndiscussed.\n", " Signals of CPT and Lorentz violation are possible in the context of\nspectroscopy using hydrogen and antihydrogen. We apply the Standard-Model\nExtension, a broad framework for Lorentz breaking in physics, to various\ntransitions in the hydrogen and antihydrogen spectra. The results show an\nunsuppressed effect in the transition between the upper two hyperfine sublevels\nof the ground state of these systems. We also discuss related tests in Penning\ntraps, and recent work on Lorentz violation in curved spacetime.\n", " We prove an adiabatic theorem for the ground state of the Dicke model in a\nslowly rotating magnetic field and show that for weak electron-photon coupling,\nthe adiabatic time scale is close to the time scale of the corresponding two\nlevel system--without the quantized radiation field. There is a correction to\nthis time scale which is the Lamb shift of the model. The photon field affect\nthe rate of approach to the adiabatic limit through a logarithmic correction\noriginating from an infrared singularity characteristic of QED.\n", " We investigate the possibility of a strongly correlated Fractional Quantum\nHall (FQH) state in bulk three dimensional isotropic (not layered) materials.\nWe find that a FQH state can exist at low densities only if it is accompanied\nby a staging transition in which the electrons re-organize themselves in\nlayers, perpendicular to the magnetic field, at distances of order the magnetic\nlength apart. The Hartree energy associated to the staging transition is\noff-set by the correlation Fock energy of the 3D FQH state. We obtain the phase\ndiagram of bulk electrons in a magnetic field subject to Coulomb interactions\nas a function of carrier density and lattice constant. At very low densities,\nthe 3D FQH state exhibits a transition to a 3D Wigner crystal state stabilized\nby phonon correlations.\n", " We investigate the thermodynamics of two-flavor quark matter in presence of\nnonzero isospin chemical potential and external magnetic field. It is known\nthat at sufficiently high isospin chemical potential, charged pions undergo\nBose-Einstein condensation (BEC). The condensate behaves as a superconductor,\nexhibiting Meissner effect in weak external magnetic fields. Stronger fields\nstress the superconducting state, turning it first into an Abrikosov lattice of\nvortices, and eventually destroying the condensate altogether. On the other\nhand, for sufficiently strong magnetic fields and low-to-moderate isospin\nchemical potential, the ground state of quantum chromodynamics (QCD) is\nexpected to be a spatially modulated condensate of neutral pions, induced by\nthe chiral anomaly: the chiral soliton lattice (CSL). We map the phase diagram\nof QCD as a function of isospin chemical potential and magnetic field in the\npart of the parameter space accessible to a low-energy effective field theory\ndescription of QCD. Our main result is an explicit account of the competition\nbetween the CSL and the Abrikosov vortex lattice. This is accomplished by\nadopting a fast numerical algorithm for finding the vortex lattice solution of\nthe equation of motion and the corresponding Gibbs energy. We find that the\nAbrikosov vortex lattice phase persists in the phase diagram, separating the\nuniform charged pion BEC phase from the CSL phase. The precise layout of the\nphase diagram depends sensitively on the choice of the vacuum pion mass.\n", " Ab initio molecular dynamics (AIMD) is a powerful tool to predict properties\nof molecular and condensed matter systems. The quality of this procedure is\nbased on accurate electronic structure calculations. The development of quantum\nprocessors has shown great potential for the efficient evaluation of accurate\nground and excited state energies of molecular systems, opening up new avenues\nfor molecular dynamics simulations. In this work we address the use of\nvariational quantum algorithms for the calculation of accurate atomic forces to\nbe used in AIMD. In particular, we provide solutions for the alleviation of the\nstatistical noise associated to the measurements of the expectation values of\nenergies and forces, as well as schemes for the mitigation of the hardware\nnoise sources (in particular, gate infidelities, qubit decoherence and readout\nerrors). Despite the relative large error in the calculation of the potential\nenergy, our results show that the proposed algorithms can provide reliable MD\ntrajectories in the microcanonical (constant energy) ensemble. Further,\nexploiting the intrinsic noise arising from the quantum measurement process, we\nalso propose a Langevin dynamics algorithm for the simulation of canonical,\ni.e., constant temperature, dynamics. Both algorithms (microcanonical and\ncanonical) are applied to the simulation of simple molecular systems such as H2\nand H3+. Finally, we also provide results for the dynamics of H2 obtained with\nIBM quantum computer ibmq_athens.\n", " The measurement of the ultra high energy cosmic ray (UHECR) spectrum is\nstrongly affected by uncertainties on the reconstructed energy. The\ndetermination of the presence or absence of the GZK cutoff and its position in\nthe energy spectrum depends not only on high statistics but also on the shape\nof the energy error distribution. Here we determine the energy error\ndistribution for fluorescence telescopes, based on a Monte Carlo simulation.\nThe HiRes and Auger fluorescence telescopes are simulated in detail. We analyze\nthe UHECR spectrum convolved with this energy error distribution. We compare\nthis spectrum with one convolved with a lognormal error distribution as well as\nwith a Gaussian error distribution. We show that the energy error distribution\nfor fluorescence detectors can not be represented by these known distributions.\nWe conclude that the convolved energy spectrum will be smeared but not enough\nto affect the GZK cutoff detection. This conclusion stands for both HiRes and\nAuger fluorescence telescopes. This result differs from the effect of the\nenergy error distribution obtained with ground detectors and reinforces the\nimportance of the fluorescence energy measurement. We also investigate the\neffect of possible fluorescence yield measurement errors in the energy\nspectrum.\n", " In this paper we study the Betti numbers of Stanley-Reisner ideals generated\nin degree 2. We show that the first six Betti numbers do not depend on the\ncharacteristic of the ground field. We also show that, if the number of\nvariables $n$ is at most 10, all Betti numbers are independent of the ground\nfield. For $n=11$, there exists precisely 4 examples in which the Betti numbers\ndepend on the ground field. This is equivalent to the statement that the\nhomology of flag complexes with at most 10 vertices is torsion free and that\nthere exists precisely 4 non-isomorphic flag complexes with 11 vertices whose\nhomology has torsion.\n In each of the 4 examples mentioned above the 8th Betti numbers depend on the\nground field and so we conclude that the highest Betti number which is always\nindependent of the ground field is either 6 or 7; if the former is true then we\nshow that there must exist a graph with 12 vertices whose 7th Betti number\ndepends on the ground field.\n", " We investigate the holographic, new agegraphic and ghost dark energy models\nin the framework of fractal cosmology. We consider a fractal FRW universe\nfilled with the dark energy and dark matter. We obtain the equation of state\nparameters of the selected dark energy models in the ultraviolet regime and\ndiscuss on their implications.\n", " We investigate the role of quantum coherence in modulating the energy\ntransfer rate between two independent energy donors and a single acceptor\nparticipating in an excitonic energy transfer process. The energy transfer rate\ndepends explicitly on the nature of the initial coherent superposition state of\nthe two donors and we connect it to the observed absorption profile of the\nacceptor and the stimulated emission profile of the energy donors. We consider\nsimple models with mesoscopic environments interacting with the donors and the\nacceptor and compare the expression we obtained for the energy transfer rate\nwith the results of numerical integration.\n", " In this work we study the first step in photosynthesis for the limiting case\nof a single photon interacting with photosystem II (PSII). We model our system\nusing quantum trajectory theory, which allows us to consider not only the\naverage evolution, but also the conditional evolution of the system given\nindividual realizations of idealized measurements of photons that have been\nabsorbed and subsequently emitted as fluorescence. The quantum nature of the\nsingle photon input requires a fully quantum model of both the input and output\nlight fields. We show that PSII coupled to the field via three collective\n``bright states'', whose orientation and distribution correlate strongly with\nits natural geometry. Measurements of the transmitted beam strongly affects the\nsystem state, since a (null) detection of the outgoing photon confirms that the\nsystem must be in the electronic (excited) ground state. Using numerical and\nanalytical calculations we show that observing the null result transforms a\nstate with a low excited state population $O( 10^{-5} )$ to a state with nearly\nall population contained in the excited states. This is solely a property of\nthe single photon input, as we confirm by comparing this behavior with that for\nexcitation by a coherent state possessing an average of one photon, using a\nsmaller five site ``pentamer'' system. We also examine the effect of a\ndissipative phononic environment on the conditional excited state dynamics. We\nshow that the environment has a strong effect on the observed rates of\nfluorescence, which could act as a new photon-counting witness of excitonic\ncoherence. The long time evolution of the phononic model predicts an\nexperimentally consistent quantum efficiency of 92%.\n", " In this study, FRW cosmology in modified gravity containing arbitrary\nfunction $f(R)$ is taken into consideration when our action are coupled with\nWeyl tensor. It is indicated that the bouncing solution emerges in the model\nwhile the equation of state (EoS) parameter crosses the phantom divider. In\nthis research, cosmological usage of the most promising candidates of dark\nenergy in the framework of $f(R)$ theory coupled by Weyl tensor is explored. A\n$f(R)$ gravity model in the spatially flat FRW universe according to the\nordinary version of the holographic dark energy model, which describes\naccelerated expansion of the universe is reconstructed. The equation of state\nparameter of the corresponding Weyl gravity models are obtained as well. We\nconclude that the holographic and Weyl gravity models can behave like phantom\nor quintessence models, whereas the equation of state parameter of the models\ncan transit from quintessence state to phantom regime as shown recent\nobservations.\n", " In the present work we have considered a modified gravity dubbed as\n\"logarithmic $f(T)$ gravity\" and investigated the behavior of Ricci dark energy\ninteracting with pressureless dark matter. We have chosen the interaction term\nin the form $Q\\propto H\\delta\\rho_{m}$ and investigated the behavior of the\nHubble parameter $H$ as a function of the redshift $z$. For this reconstructed\n$H$ we have investigated the behavior of the density of the Ricci dark energy\n$\\rho_{RDE}$ and density contribution due to torsion $\\rho_{T}$. All of the\nsaid cosmological parameters are seen to have increasing behavior from higher\nto lower redshifts for all values of $c^{2}$ pertaining to the Ricci dark\nenergy. Subsequently, we observed the equation of state parameter $w_{RDE}$ in\nthis situation. The equation of state parameter is found to behave like phantom\nfor all choices of $c^{2}$ in the Ricci dark energy.\n", " We present measurements of silica optomechanical resonators, known as bottle\nresonators, passively cooled in a cryogenic environment. These devices possess\na suite of properties that make them advantageous for preparation and\nmeasurement in the mechanical ground state, including high mechanical\nfrequency, high optical and mechanical quality factors, and optomechanical\nsideband resolution. Performing thermometry of the mechanical motion, we find\nthat the optical and mechanical modes demonstrate quantitatively similar\nlaser-induced heating, limiting the lowest average phonon occupation observed\nto just ~1500. Thermalization to the 9 mK thermal bath would facilitate quantum\nmeasurements on these promising nanogram-scale mechanical resonators.\n", " We show that the integration constant in the vacuum Schwarzschild solution\nhas the unique value, alpha_0=0, accordingly Black Holes too have the same\nunique value of mass M0=0. Therefore the so-called Black Holes Candidates (BHC)\ncannot be true BHs. It is also shown that continued collapse of sufficiently\nmassive bodies would generate radiation pressure and energy dominated\nquasistatic objects having surface gravitational redshifts z>>1. Under the\nassumption of baryon number conservation, such objects would take infinite time\nto collapse to the idealized BH state with M=0 and z=infty. The local\ntemperature of such a stellar mass Eternally Collapsing Object (ECO) would be\nabove Quark Gluon Phase transition. ECOs would undergo intermittent violent\nradiative eruptions and pollute the interstellar medium with freshly made\nhydrogen out of their QGP and also the light elements cooked in their envelope.\nIt is shown that the extremely redshifted observed temperature of an ECO could\nbe 2.75 K and superposition of ECO background radiation might generate the\nmicrowave back ground radiation. The predicted 2.75 K luminosity for the\ngalactic centre ECO, i.e., Sgr A*, L~3x1036 erg/s, nicely matches with the\ncorresponding estimate by Wilkinson Microwave Anisotropy Probe (WMAP).\n", " We consider a possibility that the formally infinite vacuum energy of the\nquantized matter fields could be stored into Planck-size quantum black holes\nacting as the fundamental constituents of space and time. Using the recently\nproposed thermodynamical equation of state obeyed by the cosmological constant\nwe indicate, how this idea might explain the smallness of the cosmological\nconstant, if the vacuum energy is the source of the cosmological constant.\n", " $\\qquad$Based on Majorana equations the e.-m. field as initially quantum\nobject having isomorphic representation as quantum field of \\guillemotleft\ndeinterlaced\\guillemotright\\ photon is considered.\n $\\qquad$The calculation of Casimir force magnitude interpreted as consequence\nof an energy measurement of the generalized quantum field of a \\guillemotleft\ndeinterlaced\\guillemotright\\ photon in the state corresponding to a\n\\guillemotleft Feynman path\\guillemotright\\ element is given. A metallic\nmirrors here plays role of classic apparatus measuring energy of this field.\n", " Using the method of QCD Sum Rules we estimate the energy of lowest energy top\nquark meson state with a hybrid admixture. This new estimate uses the previous\nestimates of hybrid charmonium and upsilon states, but with the mass of the top\nquark mass being much greater than the charm and bottom quark masses. We\ndiscuss production of mixed hybrid top quark mesons and possible detection of\nthe creation of the Quark-Gluon Plasma via Relativistic Heavy Ion Collisions.\n", " We consider the quantum entanglement of the electronic and vibrational\ndegrees of freedom in molecules with a tendency towards double welled\npotentials using model coupled harmonic diabatic potential-energy surfaces. The\nvon Neumann entropy of the reduced density matrix is used to quantify the\nelectron-vibration entanglement for the lowest two vibronic wavefunctions in\nsuch a bipartite system. Significant entanglement is found only in the region\nin which the ground vibronic state contains a density profile that is bimodal\n(i.e., contains two separate local minima). However, in this region two\ndistinct types of entanglement are found: (1) entanglement that arises purely\nfrom the degeneracy of energy levels in the two potential wells and which is\ndestroyed by slight asymmetry, and (2) entanglement that involves strongly\ninteracting states in each well that is relatively insensitive to asymmetry.\nThese two distinct regions are termed fragile degeneracy-induced entanglement\nand persistent entanglement, respectively. Six classic molecular systems\ndescribable by two diabatic states are considered: ammonia, benzene,\nsemibullvalene, pyridine excited triplet states, the Creutz-Taube ion, and the\nradical cation of the \"special pair\" of chlorophylls involved in\nphotosynthesis. These chemically diverse systems are all treated using the same\ngeneral formalism and the nature of the entanglement that they embody is\nelucidated.\n", " The relativistic quantum field theory is the unique theory that combines the\nrelativity and quantum theory and is invariant under the Poincar\\'e\ntransformation. The ground state, vacuum, is singlet and one particle states\nare transformed as elements of irreducible representation of the group. The\ncovariant one particles are momentum eigenstates expressed by plane waves and\nextended in space. Although the S-matrix defined with initial and final states\nof these states hold the symmetries and are applied to isolated states,\nout-going states for the amplitude of the event that they are detected at a\nfinite-time interval T in experiments are expressed by microscopic states that\nthey interact with, and are surrounded by matters in detectors and are not\nplane waves. These matter-induced effects modify the probabilities observed in\nrealistic situations. The transition amplitudes and probabilities of the events\nare studied with the S-matrix, $S[\\text T]$, that satisfies the boundary\ncondition at T. Using $S[\\text T]$, the finite-size corrections of the form of\n${1/\\text T}$ are found. The corrections to the Fermi's golden rule become\nlarger than the original values in some situations for light particles. They\nbreak Lorentz invariance even in high energy region of short de Broglie wave\nlengths.\n", " We study the propagation of energy density perturbation in a hot, ideal\nquark-gluon medium in which quarks and gluons follow the Tsallis-like momentum\ndistributions. We have observed that a non-extensive MIT bag equation of state\nobtained with the help of the quantum Tsallis-like distributions gives rise to\na breaking wave solution of the equation dictating the evolution of energy\ndensity perturbation. However, the breaking of waves is delayed when the value\nof the Tsallis q parameter and the Tsallis temperature T are higher.\n", " The masses of the ground state and excited heavy tetraquarks with hidden\ncharm and bottom are calculated within the relativistic diquark-antidiquark\npicture. The dynamics of the light quark in a heavy-light diquark is treated\ncompletely relativistically. The diquark structure is taken into account by\ncalculating the diquark-gluon form factor. New experimental data on\ncharmonium-like states above the open charm threshold are discussed. The\nobtained results indicate that X(3872), Y(4260), Y(4360), Z(4433) and Y(4660)\ncan be tetraquark states with hidden charm.\n", " We investigate the sign problem of the fermion determinant at finite baryon\ndensity in (1+1) dimensions, in which the ground state in the chiral limit\nshould be free from the sign problem by forming a chiral spiral. To confirm it,\nwe evaluate the fermion determinant in the continuum theory at the one-loop\nlevel and find that the determinant becomes real as expected. The conventional\nlattice formulation to implement a chemical potential is, however, not\ncompatible with the spiral transformation. We discuss an alternative of the\nfinite-density formulation and numerically verify the chiral spiral on the\nfinite-density lattice.\n", " The purpose of this contribution is to show how a nuclear field theory\nfollows naturally from the structure of four-dimensional Riemannian geometry. A\nYang-Mills field is introduced by constructing fibres that include all possible\nexchanges of spin, parity and charge such that the collective quantum numbers\nremain the same. In this way O(4) internal symmetry transformations are found\nand a connection is obtained by exponentiation of a CP-invariant operator C\nassociated with the ground state. The metric is Calabi-Yau and Einstein. Carbon\n13 is chosen as an example because it is the lightest nucleus to exhibit small\nspin mutations even though there is no deformation parameter in the O(4)\ncommutation relations. Instead a supersymmetric transformation replaces a\nquantum group. Mirror symmetry is also discussed.\n", " A novel electronic film solar energy concentrator with embedded micro-mirrors\nthat track the sun is described. The potential viability of this new concept is\npresented. Due to miniaturization, the amount of material needed for the\noptical system is minimal. Because it is light-weight and flexible, it can\neasily be attached to the land or existing structures. This presents an\neconomic advantage over conventional concentrators which require the\nconstruction of a separate structure to support them, and motors to orient them\nto intercept and properly reflect sunlight. Such separate structures must be\nable to survive gusts, windstorms, earthquakes, etc. This concentrator utilizes\nthe ground or existing edifices which are already capable of withstanding such\nvicissitudes of nature.\n", " The matrix element method (MEM) has been extensively used for the analysis of\ntop-quark and W-boson physics at the Tevatron, but in general without dedicated\ntreatment of initial state QCD radiation. At the LHC, the increased center of\nmass energy leads to a significant increase in the amount of QCD radiation,\nwhich makes it mandatory to carefully account for its effects. We here present\nseveral methods for inclusion of QCD radiation effects in the MEM, and apply\nthem to mass determination in the presence of multiple invisible particles in\nthe final state. We demonstrate significantly improved results compared to the\nstandard treatment.\n", " In this paper, a theory of dark energy is proposed that matches dark matter.\nThe relativistic quantum mechanics equations reveal that free particles can\nhave negative energies. We think that the negative energy is the dark energy\nwhich behaviors as dark photons with negative energies. In this work, the\nphoton number states are extended to the cases where the photon number can be\nnegative integers, called negative integer photon states, the physical meaning\nof which are that the photons in such a state are of negative energy, i.e.,\ndark photons. The dark photons constitute dark radiation, also called negative\nradiation. The formulism of the statistical mechanics and thermodynamics of the\ndark radiation is presented. This version of dark energy is of negative\ntemperature and negative pressure, the latter regarded as responsible for the\naccelerate expansion of the universe. It is believed that there is a symmetry\nof energy-dark energy in the universe. In our previous work, the theory of the\nmotion of the matters with negative kinetic energy was presented. In our\nopinion, the negative kinetic energy matter is dark matter. In the present\nwork, we demonstrate that the dark substances absorb and release dark energy.\nIn this view, the dark matter and dark energy match. Therefore, there is a\nsymmetry of matter-energy match and dark matter-dark energy match in the\nuniverse. We present the reasons why the negative kinetic energy systems and\nnegative radiation are dark to us.\n", " Artificial Spin Ice (ASI), consisting of a two dimensional array of nanoscale\nmagnetic elements, provides a fascinating opportunity to observe the physics of\nout of equilibrium systems. Initial studies concentrated on the static, frozen\nstate, whilst more recent studies have accessed the out-of-equilibrium dynamic,\nfluctuating state. This opens up exciting possibilities such as the observation\nof systems exploring their energy landscape through monopole quasiparticle\ncreation, potentially leading to ASI magnetricity, and to directly observe\nunconventional phase transitions. In this work we have measured and analysed\nthe magnetic relaxation of thermally active ASI systems by means of SQUID\nmagnetometry. We have investigated the effect of the interaction strength on\nthe magnetization dynamics at different temperatures in the range where the\nnanomagnets are thermally active and have observed that they follow an\nArrhenius-type N\\'eel-Brown behaviour. An unexpected negative correlation of\nthe average blocking temperature with the interaction strength is also\nobserved, which is supported by Monte Carlo simulations. The magnetization\nrelaxation measurements show faster relaxation for more strongly coupled\nnanoelements with similar dimensions. The analysis of the stretching exponents\nobtained from the measurements suggest 1-D chain-like magnetization dynamics.\nThis indicates that the nature of the interactions between nanoelements lowers\nthe dimensionality of the ASI from 2-D to 1-D. Finally, we present a way to\nquantify the effective interaction energy of a square ASI system, and compare\nit to the interaction energy calculated from a simple dipole model and also to\nthe magnetostatic energy computed with micromagnetic simulations.\n", " We show that despite the fundamentally different situations, the wave\nfunctional of the vacuum in a resonator is identical to that of free space. The\ninfinite product of Gaussian ground state wave functions defining the wave\nfunctional of the vacuum translates into an exponential of a sum rather than an\nintegral over the squares of mode amplitudes weighted by the mode volume and a\npower of the mode wave number. We express this sum by an integral of a bilinear\nform of the field containing a kernel given by a function of the square root of\nthe negative Laplacian acting on a transverse delta function. For transverse\nfields it suffices to employ the familiar delta function which allows us to\nobtain explicit expressions for the kernels of the vector potential, the\nelectric field and the magnetic induction. We show for the example of the\nvector potential that different mode expansions lead to different kernels.\nLastly, we show that the kernels have a close relationship with the Wightman\ncorrelation functions of the fields.\n" ]

[ " Ab initio QED calculations of the nuclear magnetic shielding constant in\nhelium-like ions are presented. We combine the nonrelativistic QED approach\nbased on an expansion in powers of the fine-structure constant $\\alpha$ and the\nso-called ``all-order'' QED approach which includes all orders in the parameter\n$Z\\alpha$ but uses a perturbation expansion in the parameter $1/Z$ (where $Z$\nis the nuclear charge number). The combination of the two complementary methods\nmakes our treatment applicable both to low-$Z$ and high-$Z$ ions. Our\ncalculations confirm the presence of a rare antiscreening effect for the\nrelativistic shielding correction and demonstrate the importance of the\ninclusion of the negative-energy part of the Dirac spectrum.\n", " Many NP-complete graph problems are polynomial-time solvable on graph classes\nof bounded clique-width. Several of these problems are polynomial-time solvable\non a hereditary graph class ${\\cal G}$ if they are so on the atoms (graphs with\nno clique cut-set) of ${\\cal G}$. Hence, we initiate a systematic study into\nboundedness of clique-width of atoms of hereditary graph classes. A graph $G$\nis $H$-free if $H$ is not an induced subgraph of $G$, and it is\n$(H_1,H_2)$-free if it is both $H_1$-free and $H_2$-free. A class of $H$-free\ngraphs has bounded clique-width if and only if its atoms have this property.\nThis is no longer true for $(H_1,H_2)$-free graphs, as evidenced by one known\nexample. We prove the existence of another such pair $(H_1,H_2)$ and classify\nthe boundedness of clique-width on $(H_1,H_2)$-free atoms for all but 18 cases.\n", " A graph is $(H_1,H_2)$-free for a pair of graphs $H_1,H_2$ if it contains no\ninduced subgraph isomorphic to $H_1$ or $H_2$. In 2001, Kr\\'al',\nKratochv\\'{\\i}l, Tuza, and Woeginger initiated a study into the complexity of\nColouring for $(H_1,H_2)$-free graphs. Since then, others have tried to\ncomplete their study, but many cases remain open. We focus on those\n$(H_1,H_2)$-free graphs where $H_2$ is $\\overline{H_1}$, the complement of\n$H_1$. As these classes are closed under complementation, the computational\ncomplexities of Colouring and Clique Cover coincide. By combining new and known\nresults, we are able to classify the complexity of Colouring and Clique Cover\nfor $(H,\\overline{H})$-free graphs for all cases except when $H=sP_1+ P_3$ for\n$s\\geq 3$ or $H=sP_1+P_4$ for $s\\geq 2$. We also classify the complexity of\nColouring on graph classes characterized by forbidding a finite number of\nself-complementary induced subgraphs, and we initiate a study of $k$-Colouring\nfor $(P_r,\\overline{P_r})$-free graphs.\n", " We present elastic and inelastic spin-changing cross sections for cold and\nultracold NH($X\\,^3\\Sigma^-$) + NH($X\\,^3\\Sigma^-$) collisions, obtained from\nfull quantum scattering calculations on an accurate \\textit{ab initio} quintet\npotential-energy surface. Although we consider only collisions in zero field,\nwe focus on the cross sections relevant for magnetic trapping experiments. It\nis shown that evaporative cooling of both fermionic $^{14}$NH and bosonic\n$^{15}$NH is likely to be successful for hyperfine states that allow for s-wave\ncollisions. The calculated cross sections are very sensitive to the details of\nthe interaction potential, due to the presence of (quasi-)bound state\nresonances. The remaining inaccuracy of the \\textit{ab initio} potential-energy\nsurface therefore gives rise to an uncertainty in the numerical cross-section\nvalues. However, based on a sampling of the uncertainty range of the \\textit{ab\ninitio} calculations, we conclude that the exact potential is likely to be such\nthat the elastic-to-inelastic cross-section ratio is sufficiently large to\nachieve efficient evaporative cooling. This likelihood is only weakly dependent\non the size of the channel basis set used in the scattering calculations.\n", " Recently the CIBER experiment measured the diffuse cosmic infrared background\n(CIB) flux and claimed an excess compared with integrated emission from\ngalaxies. We show that the CIB spectrum can be fitted by the additional photons\nproduced by the decay of a new particle. However, it also contributes too much\nto the anisotropy of the CIB, which is in contradiction with the anisotropy\nmeasurements by the CIBER and Hubble Space Telescope.\n", " We perform a systematic comparison of three different jet energy-loss\napproaches. These include the Armesto-Salgado-Wiedemann scheme based on the\napproach of Baier-Dokshitzer-Mueller-Peigne-Schiff and Zakharov (BDMPS-Z/ASW),\nthe Higher Twist approach (HT) and a scheme based on the approach of\nArnold-Moore-Yaffe (AMY). In this comparison, an identical medium evolution\nwill be utilized for all three approaches: not only does this entail the use of\nthe same realistic three-dimensional relativistic fluid dynamics (RFD)\nsimulation, but also includes the use of identical initial parton-distribution\nfunctions and final fragmentation functions. We are, thus, in a unique\nposition, not only to isolate fundamental differences between the various\napproaches, but also to make rigorous calculations for different experimental\nmeasurements using \"state of the art\" components. All three approaches are\nreduced to a version which contains only one free tunable parameter, this is\nthen related to the well known transport parameter $\\hat{q}$. We find that the\nparameters of all three calculations can be adjusted to provide a good\ndescription of inclusive data on $R_{AA}$ versus transverse momentum. However,\nwe do observe slight differences in their predictions for the centrality and\nazimuthal angular dependence of $R_{AA}$ vs. $p_T$. We also note that the value\nof the transport coefficient $\\hat{q}$ in the three approaches to describe the\ndata differs significantly.\n", " The angle-resolved inner-shell photoionization of R-trifluoromethyloxirane,\nC3H3F3O, is studied experimentally and theoretically. Thereby, we investigate\nthe photoelectron circular dichroism (PECD) for nearly-symmetric O 1s and F 1s\nelectronic orbitals, which are localized on different molecular sites. The\nrespective dichroic $\\beta_{1}$ and angular distribution $\\beta_{2}$ parameters\nare measured at the photoelectron kinetic energies from 1 to 16 eV by using\nvariably polarized synchrotron radiation and velocity map imaging spectroscopy.\nThe present experimental results are in good agreement with the outcome of ab\ninitio electronic structure calculations. We report a sizable chiral asymmetry\n$\\beta_{1}$ of up to about 9% for the K-shell photoionization of oxygen atom.\nFor the individual fluorine atoms, the present calculations predict asymmetries\nof similar size. However, being averaged over all fluorine atoms, it drops down\nto about 2%, as also observed in the present experiment. Our study demonstrates\na strong emitter- and site-sensitivity of PECD in the one-photon inner-shell\nionization of this chiral molecule.\n", " Dilepton production in heavy-ion collisions at top SPS energy is investigated\nwithin a coarse-graining approach that combines an underlying microscopic\nevolution of the nuclear reaction with the application of medium-modified\nspectral functions. Extracting local energy and baryon density for a grid of\nsmall space-time cells and going to each cell's rest frame enables to determine\nlocal temperature and chemical potential by application of an equation of\nstate. This allows for the calculation of thermal dilepton emission. We apply\nand compare two different spectral functions for the $\\rho$: A hadronic\nmany-body calculation and an approach that uses empirical scattering\namplitudes. Quantitatively good agreement of the model calculations with the\ndata from the NA60 collaboration is achieved for both spectral functions, but\nin detail the hadronic many-body approach leads to a better description,\nespecially of the broadening around the pole mass of the $\\rho$ and for the\nlow-mass excess. We further show that the presence of a pion chemical potential\nsignificantly influences the dilepton yield.\n", " Atomically thin two dimensional magnets have given rise to emergent phenomena\ndue to magnetic exchange and spin-orbit coupling showing a great promise for\nrealizing ultrathin device structures. In this paper, we critically examine the\nmagnetic properties of 2D FeS$_2$, which has been recently claimed to exhibit\nroom temperature ferromagnetism in the (111) orientation. Our ab initio study\nbased on collinear density functional theory has revealed the ground state as\nan antiferromagnetic one with an ordering temperature of around 100K along with\na signature of spin-phonon coupling, which may trigger a ferromagnetic coupling\nvia strain. Moreover, our calculations based on spin-spirals indicate the\npossibility of non-collinear magnetic structures, which is also supported by\nMonte Carlo simulations based on ab initio magnetic exchange parameters. This\nopens up an excellent possibility to manipulate magnetic structures by the\napplication of directional strain.\n", " We have detected, by using stimulated emission, an Atomic Parity Violation\n(APV) in the form of a chiral optical gain of a cesium vapor on the 7S -\n6P$_{3/2}$ transition,consecutive to linearly polarized 6S-7S excitation. We\ndemonstrate the validity of this detection method of APV, by presenting a 9%\naccurate measurement of expected sign and magnitude. We underline several\nadvantages of this entirely new approach in which the cylindrical symmetry of\nthe set-up can be fully exploited. Future measurements at the percent level\nwill provide an important cross-check of an existing more precise result\nobtained by a different method.\n", " Ab initio molecular dynamics (AIMD) is a powerful tool to predict properties\nof molecular and condensed matter systems. The quality of this procedure is\nbased on accurate electronic structure calculations. The development of quantum\nprocessors has shown great potential for the efficient evaluation of accurate\nground and excited state energies of molecular systems, opening up new avenues\nfor molecular dynamics simulations. In this work we address the use of\nvariational quantum algorithms for the calculation of accurate atomic forces to\nbe used in AIMD. In particular, we provide solutions for the alleviation of the\nstatistical noise associated to the measurements of the expectation values of\nenergies and forces, as well as schemes for the mitigation of the hardware\nnoise sources (in particular, gate infidelities, qubit decoherence and readout\nerrors). Despite the relative large error in the calculation of the potential\nenergy, our results show that the proposed algorithms can provide reliable MD\ntrajectories in the microcanonical (constant energy) ensemble. Further,\nexploiting the intrinsic noise arising from the quantum measurement process, we\nalso propose a Langevin dynamics algorithm for the simulation of canonical,\ni.e., constant temperature, dynamics. Both algorithms (microcanonical and\ncanonical) are applied to the simulation of simple molecular systems such as H2\nand H3+. Finally, we also provide results for the dynamics of H2 obtained with\nIBM quantum computer ibmq_athens.\n", " We describe a coupled cluster framework for coupled systems of electrons and\nphonons. Neutral and charged excitations are accessed via the\nequation-of-motion version of the theory. Benchmarks on the Hubbard-Holstein\nmodel allow us to assess the strengths and weaknesses of different coupled\ncluster approximations which generally perform well for weak to moderate\ncoupling. Finally, we report progress towards an implementation for {\\it ab\ninitio} calculations on solids, and present some preliminary results on\nfinite-size models of diamond. We also report the implementation of\nelectron-phonon coupling matrix elements from crystalline Gaussian type\norbitals (cGTO) within the PySCF program package.\n", " In this work, we have continued the study of the Hawking radiation on the\nbrane from a higher-dimensional rotating black hole by investigating the\nemission of fermionic modes. A comprehensive analysis is performed that leads\nto the particle, power and angular momentum emission rates, and sheds light on\ntheir dependence on fundamental parameters of the theory, such as the spacetime\ndimension and angular momentum of the black hole. In addition, the angular\ndistribution of the emitted modes, in terms of the number of particles and\nenergy, is thoroughly studied. Our results are valid for arbitrary values of\nthe energy of the emitted particles, dimension of spacetime and angular\nmomentum of the black hole, and complement previous results on the emission of\nbrane-localised scalars and gauge bosons.\n", " We study hydrogenated cations of two polycyclic hydrocarbon molecules as\nmodels of hydrogenated organic species that form in the interstellar medium.\nOptical spectra of the hydrogenated naphthalene cation Hn-C10H8^+ for n=1,2,\nand 10, as well as of the astrobiologically interesting hydrogenated proflavine\ncation Hn-C13H11N3^+ for n=1 and 14, are calculated. The pseudopotential time\ndependent density functional theory is used. It is found that the fully\nhydrogenated proflavine cation H14-C13H11N3^+ shows a broad spectrum where\npositions of individual lines are almost lost. The positions of lines, their\nshapes, and intensities change in hydronaphthalene and hydroproflavine cations\nshowing that hydrogen additions induce substantially different optical spectra\nin a comparison to base polycyclic hydrocarbon cations. One calculated line in\nthe visible spectrum of H10-C10H8^+, and one in the visible spectrum of\nH-C13H11N3^+ are close to the measured diffuse interstellar bands. We also\npresent positions of near-UV lines.\n", " Let $H_1$, $H_2$ be complex Hilbert spaces. A bounded linear operator $T :\nH_1 \\to H_2$ is said to be norm attaining if there exists a unit vector $x \\in\nH_1$ such that $\\|Tx\\| = \\|T\\|$. If $T|_{M} : M \\to H_2$ is norm attaining for\nevery closed subspace $M$ of $H_1$, then we say that $T$ is an absolutely norm\nattaining ($\\mathcal{AN}$-operator). If the norm of the operator is replaced by\nthe minimum modulus $m(T) = \\inf\\{\\|Tx\\| : x \\in H_1, \\|x\\| =1\\}$, then $T$ is\nsaid to be a minimum attaining and an absolutely minimum attaining operator\n($\\mathcal{AM}$-operator), respectively.\n In this article, we give representations of quasinormal $\\mathcal{AN}$,\n$\\mathcal{AM}$-operators and the operators in the closure of these two classes.\nLater we extend these results to the class of hyponormal operators in the\nclosure of $\\mathcal{AN}$-operators and a further look at some sufficient\nconditions under which these operators become normal.\n", " When monoclinic monazite-type LaVO4 (space group P21/n) is squeezed up to 12\nGPa at room temperature, a phase transition to another monoclinic phase has\nbeen found. The structure of the high-pressure phase of LaVO4 is indexed with\nthe same space group (P21/n), but with a larger unit-cell in which the number\nof atoms is doubled. The transition leads to an 8% increase in the density of\nLaVO4. The occurrence of such a transition has been determined by x-ray\ndiffraction, Raman spectroscopy, and ab initio calculations. The combination of\nthe three techniques allows us to also characterize accurately the pressure\nevolution of unit-cell parameters and the Raman (and IR)-active phonons of the\nlow- and high-pressure phase. In particular, room-temperature equations of\nstate have been determined. The changes driven by pressure in the crystal\nstructure induce sharp modifications in the color of LaVO4 crystals, suggesting\nthat behind the monoclinic-to-monoclinic transition there are important changes\nof the electronic properties of LaVO4.\n", " Fragmentation functions are determined for the pion and kaon by global\nanalyses of charged-hadron production data in electron-positron annihilation.\nAccurate measurements were reported by the Belle and BaBar collaborations for\nthe fragmentation functions at the center-of-mass energies of 10.52 GeV and\n10.54 GeV, respectively, at the KEK and SLAC B factories, whereas other\navailable $e^+e^-$ measurements were mostly done at higher energies, mainly at\nthe $Z$ mass of 91.2 GeV. There is a possibility that gluon fragmentation\nfunctions, as well as quark fragmentation functions, are accurately determined\nby scaling violation. We report our global analysis of the fragmentation\nfunctions especially to show impacts of the B-factory measurements on the\nfragmentation function determination. Our results indicate that the\nfragmentation functions are determined more accurately not only by the scaling\nviolation but also by high-statistical nature of the Belle and BaBar data.\nHowever, there are some tensions between the Belle and BaBar data in comparison\nwith previous measurements. We also explain how the flavor dependence of quark\nfragmentation functions and the gluon function are separated by using\nmeasurements at different $Q^2$ values. In particular, the electric and weak\nchanges are different depending on the quark type, so that a light-quark flavor\nseparation also became possible in principle due to the precise data at both\n$\\sqrt{s}\\simeq 10.5$ GeV and 91.2 GeV.\n", " Within the next-to-leading order (NLO) perturbative QCD (pQCD) parton model,\nsuppression of away-side hadron spectra associated with a high $p_{T}$ photon\ndue to parton energy loss is shown to provide a complete tomographic picture of\nthe dense matter formed in high-energy heavy-ion collisions at RHIC. Dictated\nby the shape of the $\\gamma$-triggered jet spectrum in NLO pQCD, hadron spectra\nat large $z_T=p_T^{h}/p_T^{\\gamma} \\stackrel{>}{\\sim} 1$ are more susceptible\nto parton energy loss and therefore are dominated by surface emission of\n$\\gamma$-triggered jets, whereas small $z_{T}$ hadrons mainly come from\nfragmentation of jets with reduced energy from volume emission. These lead to\ndifferent centrality dependence of the hadron suppression in different regions\nof $z_{T}$.\n", " The use of QCD calculations that include the resummation of soft-collinear\nlogarithms via parton-shower algorithms is currently not possible in PDF fits\ndue to the high computational cost of evaluating observables for each variation\nof the PDFs. Unfortunately the interpolation methods that are otherwise applied\nto overcome this issue are not readily generalised to all-order parton-shower\ncontributions. Instead, we propose an approximation based on training a neural\nnetwork to predict the effect of varying the input parameters of a parton\nshower on the cross section in a given observable bin, interpolating between\nthe variations of a training data set. This first publication focuses on\nproviding a proof-of-principle for the method, by varying the shower dependence\non $\\alpha_\\text{S}$ for both a simplified shower model and a complete shower\nimplementation for three different observables, the leading emission scale, the\nnumber of emissions and the Thrust event shape. The extension to the PDF\ndependence of the initial-state shower evolution that is needed for the\napplication to PDF fits is left to a forthcoming publication.\n", " Nonadiabatic corrections in molecules composed of a few atoms are considered.\nIt is demonstrated that a systematic perturbative expansion around the\nadiabatic solution is possible, with the expansion parameter being the\nelectron-nucleus mass ratio to the power 3/4. Closed form formulae for the\nleading corrections to the wave function and to the energy are derived. Their\napplicability is demonstrated by a comparison of numerical results for the\nhydrogen molecule with the former nonadiabatic calculations and the\nexperimental values. Good agreement with the recent experiment is achieved for\nthe ground state dissociation energy of both H$_2$ and D$_2$.\n", " We study the $\\bar B^0_s \\to J/\\psi K^+ K^-$, $\\bar B^0 \\to J/\\psi K^+ K^-$,\n$B^- \\to J/\\psi K^0 K^-$, $\\bar B^0 \\to J/\\psi \\pi^0 \\eta$ and $B^- \\to J/\\psi\n\\pi^- \\eta$ decays and compare their mass distributions with those obtained for\nthe $\\bar B^0_s \\to J/\\psi \\pi^+ \\pi^-$ and $\\bar B^0 \\to J/\\psi \\pi^+ \\pi^-$.\nThe approach followed consist in a factorization of the weak part and the\nhadronization part into a factor which is common to all the processes. Then\nwhat makes the reactions different are some trivial Cabibbo-Kobayashi-Maskawa\nmatrix elements and the weight by which the different pairs of mesons appear in\na primary step plus their final state interaction. These elements are part of\nthe theory and thus, up to a global normalization factor, all the invariant\nmass distributions are predicted with no free parameters. Comparison is made\nwith the limited experimental information available. Further comparison of\nthese results with coming LHCb measurements will be very valuable to make\nprogress in our understanding of the meson-meson interaction and the nature of\nthe low lying scalar meson resonances, $f_0(500), f_0(980)$ and $a_0(980)$.\n", " New calculations for rotational excitation of cyanoacetylene by collisions\nwith hydrogen molecules are performed to include the lowest 38 rotational\nlevels of HC3N and kinetic temperatures to 300 K. Calculations are based on the\ninteraction potential of Wernli et al. A&A, 464, 1147 (2007) whose accuracy is\nchecked against spectroscopic measurements of the HC3N-H2 complex. The quantum\ncoupled-channel approach is employed and complemented by quasi-classical\ntrajectory calculations. Rate coefficients for ortho-H2 are provided for the\nfirst time. Hyperfine resolved rate coefficients are also deduced. Collisional\npropensity rules are discussed and comparisons between quantum and classical\nrate coefficients are presented. This collisional data should prove useful in\ninterpreting HC3N observations in the cold and warm ISM, as well as in\nprotoplanetary disks.\n", " The spectral properties of a singular left-definite Sturm-Liouville operator\n$JA$ are investigated and described via the properties of the corresponding\nright-definite selfadjoint counterpart $A$ which is obtained by substituting\nthe indefinite weight function by its absolute value. The spectrum of the\n$J$-selfadjoint operator $JA$ is real and it follows that an interval\n$(a,b)\\subset\\mathbb R^+$ is a gap in the essential spectrum of $A$ if and only\nif both intervals $(-b,-a)$ and $(a,b)$ are gaps in the essential spectrum of\nthe $J$-selfadjoint operator $JA$. As one of the main results it is shown that\nthe number of eigenvalues of $JA$ in $(-b,-a) \\cup (a,b)$ differs at most by\nthree of the number of eigenvalues of $A$ in the gap $(a,b)$; as a byproduct\nresults on the accumulation of eigenvalues of singular left-definite\nSturm-Liouville operators are obtained. Furthermore, left-definite problems\nwith symmetric and periodic coefficients are treated, and several examples are\nincluded to illustrate the general results.\n", " This chapter provides a comprehensive review of fundamental concepts related\nto approximate natural orbital functionals (NOFs), emphasizing their\nsignificance in quantum chemistry and physics. Focusing on fermions, the\ndiscussion excludes considerations of finite temperature and systems with a\nvariable number of particles. The theoretical foundation for approximate NOFs\nis laid out, with a particular emphasis on functional N-representability.\nVarious two-index reconstructions for the two-particle reduced density matrix\n(2RDM) are introduced, accompanied by discussions on challenges. The analysis\ndelves deeply into NOFs grounded in electron pairing, specifically focusing on\nPNOF5, PNOF7, and the Global NOF, a more versatile approach addressing both\nstatic and dynamic electron correlation components. The extension of NOFs to\nmultiplets while conserving total spin is presented, and the availability of\nopen-source implementations like DoNOF (http://github.com/DoNOF) and its\nassociated programs is highlighted. A detailed overview of optimization\nprocedures for single-point calculations is provided. Sections on geometry\noptimization and ab initio molecular dynamics, closely connected to the\navailability of analytical gradients in NOF theory, are presented. The chapter\nconcludes with the extension of NOFs to both charged and excited states.\n", " The availability of the double-differential charged-current neutrino cross\nsection, measured by the MiniBooNE collaboration using a carbon target, allows\nfor a systematic comparison of nuclear effects in quasi-elastic electron and\nneutrino scattering. The results of our study, based on the impulse\napproximation scheme and a state-of-the-art model of the nuclear spectral\nfunctions, suggest that the electron cross section and the flux averaged\nneutrino cross sections, corresponding to the same target and comparable\nkinematical conditions, cannot be described within the same theoretical\napproach using the value of the nucleon axial mass obtained from deuterium\nmeasurements. We analyze the assumptions underlying the treatment of electron\nscattering data, and argue that the description of neutrino data will require a\nnew {\\em paradigm}, suitable for application to processes in which the lepton\nkinematics is not fully determined.\n", " We review some of the results obtained in the study of the production of\nelectromagnetic radiation in relativistic nuclear collisions. We concentrate on\nthe emission of real photons and dileptons from the hot and dense strongly\ninteracting phases of the reaction. We examine the contributions from the\npartonic sector, as well as those from the nonperturbative hadronic sector. We\nexamine the current data, some of the predictions for future measurements, and\ncomment on what has been learnt so far.\n", " Models of our galaxy based on dynamical observations predict a spheroid\ncomponent much heavier than accounted for by direct measurements of star counts\nand high velocity stars. If, as first suggested by Caldwell and Ostriker, this\ndiscrepancy is due to a large population of faint low-mass stars or dark\nobjects in the spheroid, the spheroid could be responsible for microlensing\nevents for sources in the Large Magellanic Cloud (LMC). We show that, although\nthe rate of events is lower than predicted by a galactic halo made of\nmicrolensing objects, it is still significant for EROS/MACHO observations.\nBecause of the different matter distributions in the halo and spheroid\ncomponents, a comparison between microlensing event rates in the LMC, future\nmeasurements of microlensing in the galactic bulge and, possibly, in M31 can\nprovide information about the amounts of dark objects in the different galactic\ncomponents. If the EROS/MACHO collaborations find a deficiency with respect to\ntheir halo expectation, when more statistics are available, their detected\nevents could be interpreted as coming from spheroid microlenses, allowing for a\ngalactic halo composed entirely of non-baryonic dark matter.\n", " The usual procedure for estimating the significance of a peak in a power\nspectrum is to calculate the probability of obtaining that value or a larger\nvalue by chance, on the assumption that the time series contains only noise\n(e.g. that the measurements were derived from random samplings of a Gaussian\ndistribution). However, it is known that one should regard this P-Value\napproach with caution. As an alternative, we here examine a Bayesian approach\nto estimating the significance of a peak in a power spectrum. This approach\nrequires that we consider explicitly the hypothesis that the time series\ncontains a periodic signal as well as noise. The challenge is to identify a\nprobability distribution function for the power that is appropriate for this\nhypothesis. We propose what seem to be reasonable conditions to require of this\nfunction, and then propose a simple function that meets these requirements. We\nalso propose a consistency condition, and check to see that our function\nsatisfies this condition. We find that the Bayesian significance estimates are\nconsiderably more conservative than the conventional estimates. We apply this\nprocedure to three recent analyses of solar neutrino data: (a) bimodality of\nGALLEX data; (b) power spectrum analysis of Super-Kamiokande data; and (c) the\ncombined analysis of radiochemical neutrino data and irradiance data.\n", " The results of density functional theory calculations and measurements using\nX-ray photoelectron spectroscopy of Co-nanoparticles dispersed on graphene/Cu\nare presented. It is found that for low cobalt thickness (0.02 nm - 0.06 nm)\nthe Co forms islands distributed non-homogeneously which are strongly oxidized\nunder exposure to air to form cobalt oxides. At greater thicknesses up to 2 nm\nthe upper Co-layers are similarly oxidized whereas the lower layers contacting\nthe graphene remain metallic. The measurements indicate a Co2+ oxidation state\nwith no evidence of a 3+ state appearing at any Co thickness, consistent with\nCoO and Co[OH]2. The results show that thicker Co (2nm) coverage induces the\nformation of a protective oxide layer while providing the magnetic properties\nof Co nanoparticles.\n", " The neutrino magnetic and electric moments are zero at tree level but can\narise in radiative corrections. Any deviation from the Standard Model\nprediction would provide another indication of neutrino-related new physics in\naddition to the neutrino oscillation and masses. Especially, Dirac and Majorana\nneutrinos have quite different structures in their electromagnetic moments.\nNevertheless, the recoil measurements and astrophysical stellar cooling can\nonly constrain combinations of neutrino magnetic and electric moments with the\nlimitation of not seeing their detailed structures. We propose using the atomic\nradiative emission of neutrino pair to serve as a unique probe of the neutrino\nelectromagnetic moments with the advantage of not just separating the magnetic\nand electric moments but also identifying their individual elements. Both\nsearching strategy and projected sensitivities are illustrated in this letter.\n", " The Fermi-LAT detects no significant anisotropy of the cosmic-ray (CR)\nelectrons and positrons ($e^-+e^+$) with seven years of data, which provides\nthe strongest restriction to the $e^-+e^+$ anisotropy up to now. As next\ngeneration CR observatory, HERD is expected to have a better capability of\nanisotropy detection than Fermi-LAT. In this paper, we discuss several models\naimed to explain the AMS-02 data by the present and future anisotropy\nmeasurements. We find that the upper limits of Fermi-LAT disfavor Vela SNR as\nthe dominant source in sub-TeV, while other cases that remain safe under the\nconstraint of Fermi-LAT are expected to be distinguished from each other by\nHERD. We then discuss the possibilities of remarkable TeV spectral features,\nand test the corresponding anisotropies. We find the conditions under which the\nTeV model can have a prominent spectral feature and avoid the constraint of\nFermi-LAT at the same time. Furthermore, the expected performance of HERD is\nsensitive enough to detect the anisotropies of all these TeV models, and even\nfor the case of a featureless TeV spectrum. Thus HERD may play a crucial part\nin the study of the origin of cosmic electrons and positrons.\n", " We study the existence and linear stability of stationary periodic solutions\nto a neural field model, an intergo-differential equation of the Hammerstein\ntype. Under the assumption that the activation function is a discontinuous step\nfunction and the kernel is decaying sufficiently fast, we formulate necessary\nand sufficient conditions for the existence of a special class of solutions\nthat we call 1-bump periodic solutions. We then analyze the stability of these\nsolutions by studying the spectrum of the Frechet derivative of the\ncorresponding Hammerstein operator. We prove that the spectrum of this operator\nagrees up to zero with the spectrum of a block Laurent operator. We show that\nthe non-zero spectrum consists of only eigenvalues and obtain an analytical\nexpression for the eigenvalues and the eigenfunctions. The results are\nillustrated by multiple examples.\n", " Thermal conductivity measurements over variable lengths on nanostructures\nsuch as nanowires provide important information about the mean free paths\n(MFPs) of the phonons responsible for heat conduction. However, nearly all of\nthese measurements have been interpreted using an average MFP even though\nphonons in many crystals possess a broad MFP spectrum. Here, we present a\nreconstruction method to obtain MFP spectra of nanostructures from\nvariable-length thermal conductivity measurements. Using this method, we\ninvestigate recently reported length-dependent thermal conductivity\nmeasurements on SiGe alloy nanowires and suspended graphene ribbons. We find\nthat the recent measurements on graphene imply that 70 % of the heat in\ngraphene is carried by phonons with MFPs longer than 1 micron.\n", " In this work we study the first step in photosynthesis for the limiting case\nof a single photon interacting with photosystem II (PSII). We model our system\nusing quantum trajectory theory, which allows us to consider not only the\naverage evolution, but also the conditional evolution of the system given\nindividual realizations of idealized measurements of photons that have been\nabsorbed and subsequently emitted as fluorescence. The quantum nature of the\nsingle photon input requires a fully quantum model of both the input and output\nlight fields. We show that PSII coupled to the field via three collective\n``bright states'', whose orientation and distribution correlate strongly with\nits natural geometry. Measurements of the transmitted beam strongly affects the\nsystem state, since a (null) detection of the outgoing photon confirms that the\nsystem must be in the electronic (excited) ground state. Using numerical and\nanalytical calculations we show that observing the null result transforms a\nstate with a low excited state population $O( 10^{-5} )$ to a state with nearly\nall population contained in the excited states. This is solely a property of\nthe single photon input, as we confirm by comparing this behavior with that for\nexcitation by a coherent state possessing an average of one photon, using a\nsmaller five site ``pentamer'' system. We also examine the effect of a\ndissipative phononic environment on the conditional excited state dynamics. We\nshow that the environment has a strong effect on the observed rates of\nfluorescence, which could act as a new photon-counting witness of excitonic\ncoherence. The long time evolution of the phononic model predicts an\nexperimentally consistent quantum efficiency of 92%.\n", " The photon box thought experiment can be considered a forerunner of the\nEPR-experiment: by performing suitable measurements on the box it is possible\nto ``prepare'' the photon, long after it has escaped, in either of two\ncomplementary states. Consistency requires that the corresponding box\nmeasurements be complementary as well. At first sight it seems, however, that\nthese measurements can be jointly performed with arbitrary precision: they\npertain to different systems (the center of mass of the box and an internal\nclock, respectively). But this is deceptive. As we show by explicit\ncalculation, although the relevant quantities are simultaneously measurable,\nthey develop non-vanishing commutators when calculated back to the time of\nescape of the photon. This justifies Bohr's qualitative arguments in a precise\nway; and it illustrates how the details of the dynamics conspire to guarantee\nthe requirements of complementarity. In addition, our calculations exhibit a\n``fine structure'' in the distribution of the uncertainties over the\ncomplementary quantities: depending on \\textit{when} the box measurement is\nperformed, the resulting quantum description of the photon differs. This brings\nus close to the argumentation of the later EPR thought experiment.\n", " We present a geometrical construction of families of finite isospectral\ngraphs labelled by different partitions of a natural number $r$ of given length\n$s$ (the number of summands). Isospectrality here refers to the discrete\nmagnetic Laplacian with normalised weights (including standard weights). The\nconstruction begins with an arbitrary finite graph $G$ with normalised weight\nand magnetic potential as a building block from which we construct, in a first\nstep, a family of so-called frame graphs $(F_a)_{a \\in \\mathbb{N}}$. A frame\ngraph $F_a$ is constructed contracting $a$ copies of $G$ along a subset of\nvertices $V_0$. In a second step, for any partition $A=(a_1,\\dots,a_s)$ of\nlength $s$ of a natural number $r$ (i.e., $r=a_1+\\dots+a_s$) we construct a new\ngraph $F_A$ contracting now the frames $F_{a_1},\\dots,F_{a_s}$ selected by $A$\nalong a proper subset of vertices $V_1\\subset V_0$. All the graphs obtained by\ndifferent $s$-partitions of $r\\geq 4$ (for any choice of $V_0$ and $V_1$) are\nisospectral and non-isomorphic.\n In particular, we obtain increasing finite families of graphs which are\nisospectral for given $r$ and $s$ for different types of magnetic Laplacians\nincluding the standard Laplacian, the signless standard Laplacian, certain\nkinds of signed Laplacians and, also, for the (unbounded) Kirchhoff Laplacian\nof the underlying equilateral metric graph. The spectrum of the isospectral\ngraphs is determined by the spectrum of the Laplacian of the building block $G$\nand the spectrum for the Laplacian with Dirichlet conditions on the set of\nvertices $V_0$ and $V_1$ with multiplicities determined by the numbers $r$ and\n$s$ of the partition.\n", " We argue that certain assumptions about the process of the emission of the\nquanta by their (oscillating) emitter provide for their changing (oscillatory)\nflux at any location. This mechanism underlying (such) wave phenomena is not\nbased, both, on the newtonian notion of force and the field concept (of\nFaraday, Maxwell, Lorentz and Einstein). When applied to the case of thermal\nradiation, this emission origin for the wave of quanta is shown here to be\nconsistent with the laws of the black body radiation. We conclude therefore\nalso that a conceptual framework, which is not rooted in the notion of force\nand in the field concept, may provide a deterministic basis underlying the\nprobabilistic methods of the quantum theory.\n", " A renewed analysis of the H.E. Ives and G.R. Stilwell's experiment on moving\nhydrogen canal rays (J. Opt. Soc. Am., 1938, v.28, 215) concludes that the\nspectral emission of a moving atom exhibits always a redshift which informs not\nthe direction of the atom's motion. The conclusion is also evident from a\nsimple energy relation: atomic spectral radiation is emitted as an orbiting\nelectron consumes a portion of its internal energy on transiting to a\nlower-energy state which however has in a moving atom an additional energy\ngain; this results in a redshift in the emission frequency. Based on auxiliary\nexperimental information and a scheme for de Broglie particle formation, we\ngive a vigorous elucidation of the mechanism for deceleration radiation of\natomic electron; the corresponding prediction of the redshift is in complete\nagreement with the Ives and Stilwell's experimental formula.\n", " We propose a new fast method to match factorization theorems applicable in\ndifferent kinematical regions, such as the transverse-momentum-dependent and\nthe collinear factorization theorems in Quantum Chromodynamics. At variance\nwith well-known approaches relying on their simple addition and subsequent\nsubtraction of double-counted contributions, ours simply builds on their\nweighting using the theory uncertainties deduced from the factorization\ntheorems themselves. This allows us to estimate the unknown complete matched\ncross section from an inverse-error-weighted average. The method is simple and\nprovides an evaluation of the theoretical uncertainty of the matched cross\nsection associated with the uncertainties from the power corrections to the\nfactorization theorems (additional uncertainties, such as the nonperturbative\nones, should be added for a proper comparison with experimental data). Its\nusage is illustrated with several basic examples, such as $Z$ boson, $W$ boson,\n$H^0$ boson and Drell-Yan lepton-pair production in hadronic collisions, and\ncompared to the state-of-the-art Collins-Soper-Sterman subtraction scheme. It\nis also not limited to the transverse-momentum spectrum, and can\nstraightforwardly be extended to match any (un)polarized cross section\ndifferential in other variables, including multi-differential measurements.\n", " We present direct measurements of depth-resolved band gap variations of\nCuIn(1-x)Ga(x)Se2 thin-film solar cell absorbers. A new measurement technique\ncombining parallel measurements of local thin-film interference and spectral\nphotoluminescence was developed for this purpose. We find sample-dependent\ncorrelation parameters between measured band gap depth and composition\nprofiles, and emphasize the importance of direct measurements. These results\nbring a quantitative insight into the electronic properties of the solar cells\nand open a new way to analyze parameters that determine the efficiency of solar\ncells.\n", " We consider Mirror Twin Higgs models in which the breaking of the global\nsymmetry is realized linearly. In this scenario, the radial mode in the Higgs\npotential is present in the spectrum, and constitutes a second portal between\nthe twin and SM sectors. We show that a study of the properties of this\nparticle at colliders, when combined with precision measurements of the light\nHiggs, can be used to overdetermine the form of the scalar potential, thereby\nconfirming that it possesses an enhanced global symmetry as dictated by the\nTwin Higgs mechanism. We find that, although the reach of the LHC for this\nstate is limited, future linear colliders will be able to explore a significant\npart of the preferred parameter space, allowing the possibility of directly\ntesting the Twin Higgs framework.\n", " Positron binding energies in halogenated hydrocarbons are calculated \\emph{ab\ninitio} using many-body theory. For chlorinated molecules, including planars\nfor which the interaction is highly anisotropic, very good to excellent\nagreement with experiment and recent DFT-based model-potential calculations is\nfound. Predictions for fluorinated and brominated molecules are presented. The\ncomparative effect of fluorination, chlorination and bromination is elucidated\nby identifying trends within molecular families including dihaloethylenes and\nhalomethanes based on global molecular properties (dipole moment,\npolarizability, ionization energy). It is shown that relative to brominated and\nchlorinated molecules, fluorinated molecules generate a less attractive\npositron-molecule potential due to larger ionization energies and smaller\ndensity of molecular orbitals close to the HOMO, resulting in very weak, or in\nmost cases loss of, positron binding. Overall, however, it is shown that the\nglobal molecular properties are not universal predictors of binding energies,\nexemplified by consideration of CH$_3$Cl vs.~\\emph{cis.}-C$_2$H$_2$F$_2$:\ndespite the latter having a larger dipole moment, lower ionization energy and\nsimilar polarizability its binding energy is significantly smaller (25 meV\nvs.~3 meV, respectively), owing to the important contribution of multiple\nmolecular orbitals to, and the anisotropy of, the positron-molecule correlation\npotential.\n", " The intriguing connection between black holes' evaporation and physics of\nsolitons is opening novel roads to finding observable phenomena. It is known\nfrom the inverse scattering transform that velocity is a fundamental parameter\nin solitons theory. Taking this into account, the study of Haw\\-king radiation\nby a moving soliton gets a growing relevance. However, a theoretical context\nfor the description of this phenomenon is still lacking. Here, we adopt a\nsoliton geometrization technique to study the quantum emission of a moving\nsoliton in a one-dimensional model. Representing a black hole by the one\nsoliton solution of the sine-Gordon equation, we consider Haw\\-king emission\nspectra of a quantized massless scalar field on the soliton-induced metric. We\nstudy the relation between the soliton velocity and the black hole temperature.\nOur results address a new scenario in the detection of new physics in the\nquantum gravity panorama.\n", " Consider an elliptic self-adjoint pseudodifferential operator $A$ acting on\n$m$-columns of half-densities on a closed manifold $M$, whose principal symbol\nis assumed to have simple eigenvalues. Relying on a basis of pseudodifferential\nprojections commuting with $A$, we construct an almost-unitary\npseudodifferential operator that diagonalizes $A$ modulo an infinitely\nsmoothing operator. We provide an invariant algorithm for the computation of\nits full symbol, as well as an explicit closed formula for its subprincipal\nsymbol. Finally, we give a quantitative description of the relation between the\nspectrum of $A$ and the spectrum of its approximate diagonalization, and\ndiscuss the implications at the level of spectral asymptotics.\n", " Strong correlation can be essentially captured with multireference\nwavefunction methods such as complete active space self-consistent field\n(CASSCF) or density matrix renormalization group (DMRG). Still, an accurate\ndescription of the electronic structure of strongly correlated systems requires\naccounting for the dynamic electron correlation, which CASSCF and DMRG largely\nmiss. In this work a new approach for the correlation energy based on the\nadiabatic connection (AC) is proposed. The AC$_{\\rm n}$ method accounts for\nterms up to the desired order n in the coupling constant, is rigorously\nsize-consistent, free from instabilities and intruder states. It employs the\nparticle-hole multireference random phase approximation and the Cholesky\ndecomposition technique, which leads to a computational cost growing with the\nfifth power of the system size. Thanks to AC$_{\\rm n}$ depending solely on one-\nand two-electron CAS reduced density matrix, the method is much more efficient\nthan existing ab initio dynamic correlation methods for strong correlation.\nAC$_{\\rm n}$ affords excellent results for singlet-triplet gaps of challenging\norganic biradicals. Development presented in this work opens new perspectives\nfor accurate calculations of systems with dozens of strongly correlated\nelectrons.\n", " Technology moves towards ever faster switching between different electronic\nand magnetic states of matter. Manipulating properties at terahertz rates\nrequires accessing the intrinsic timescales of electrons (femtoseconds) and\nassociated phonons (10s of femtoseconds to few picoseconds), which is possible\nwith short-pulse photoexcitation. Yet, in many Mott insulators, the electronic\ntransition is accompanied by the nucleation and growth of percolating domains\nof the changed lattice structure, leading to empirical time scales dominated by\nslow coarsening dynamics. Here, we use time-resolved X-ray diffraction and\nreflectivity measurements to investigate the photoinduced insulator-to-metal\ntransition in an epitaxially strained thin film Mott insulator Ca2RuO4. The\ndynamical transition occurs without observable domain formation and coarsening\neffects, allowing the study of the intrinsic electronic and lattice dynamics.\nAbove a fluence threshold, the initial electronic excitation drives a fast\nlattice rearrangement, followed by a slower electronic evolution into a\nmetastable non-equilibrium state. Microscopic calculations based on\ntime-dependent dynamical mean-field theory and semiclassical lattice dynamics\nwithin a recently published equilibrium energy landscape picture explain the\nthreshold-behavior and elucidate the delayed onset of the electronic phase\ntransition in terms of kinematic constraints on recombination. Analysis of\nsatellite scattering peaks indicates the persistence of a strain-induced\nnano-texture in the photoexcited film. This work highlights the importance of\ncombined electronic and structural studies to unravel the physics of dynamic\ntransitions and elucidates the role of strain in tuning the timescales of\nphotoinduced processes.\n", " Reactions $n_{\\gamma }\\gamma +$ $d+$ $_{z_{2}}^{A_{2}}X\\rightarrow \\left(\nn_{\\gamma }\\pm 1\\right) \\gamma +$ $p+$ $_{z_{2}}^{A_{2}+1}X$ and especially the\nreaction $d+$ $_{z_{2}}^{A_{2}}X\\rightarrow \\gamma +$ $p+$ $%\n_{z_{2}}^{A_{2}+1}X$, called photo-induced nuclear cooperation and cooperative\nspontaneous $\\gamma $ emission with neutron exchange, respectively, are\ninvestigated theoretically. In the case of photo-induced nuclear cooperation it\nis supposed that the energy of $\\gamma $ photons of the beam is less than the\nbinding energy of the deuteron. The cross section and the transition\nprobability per unit time, respectively, are determined with the aid of\nstandard second order perturbation calculation of quantum mechanics. The\ncalculations are extended to photo-induced nuclear cooperation and cooperative\nspontaneous $\\gamma $ emission with proton exchange as well. With the aid of\nthe results obtained, recent observations of nuclear activity of samples of\nlarge deuteron content after irradiation by photon-flux of photon energy\nsmaller than the deuteron binding energy are discussed.\n", " The burgeoning interest in 2D black phosphorus (bP) contributes to expand its\napplications in countless fields. In the present study, 2D bP is used as a\nsupport for homogeneously dispersed palladium nanoparticles directly grown on\nit by a wet chemical process. EELS-STEM analysis evidences a strong interaction\nbetween palladium and P atoms of bP nanosheets. A quantitative evaluation of\nthis interaction comes from XAS measurements that find out a very short Pd-P\ndistance of 2.26 {\\AA} proving for the first time the existence of an\nunprecedented Pd-P coordination bond of covalent nature. Additionally, the\naverage Pd-P coordination number of about 1.7 reveals that bP acts as a\npolydentate phosphine ligand towards the surface Pd atoms of the nanoparticles,\nthus preventing their agglomeration and inferring structural stability. These\nunique properties result in a superior performance in the catalytic\nhydrogenation of chloronitroarenes to chloroaniline, where a higher\nchemoselectivity in comparison to other heterogeneous catalyst based on\npalladium has been observed.\n", " We have given theoretical expressions for the forces exerted on a quasi-2D\nflat and smooth solid plate immersed into a liquid pool of a simple liquid. All\nforces given by the theory, the local forces on the top, the contact line and\nthe bottom of the plate as well as the total force, showed an excellent\nagreement with the MD simulation results. We also revealed that the local\nforces around the bottom and top of the solid plate can be related to the SL\nand SV interfacial tensions \\gamma_{SL} and \\gamma_{SV}, and this was verified\nthrough the comparison with the SL and SV works of adhesion obtained by the\nthermodynamic integration. We have reached a conclusion that \\gamma_{SL} and\n\\gamma_{SV} as well as the liquid-vapor interfacial tension \\gamma_{LV} can be\nextracted from a single equilibrium MD simulation without the\ncomputationally-demanding calculations.\n", " We consider emission of a photon by an electron in the field of a strong\nlaser wave. A probability of this process for circularly or linearly polarized\nlaser photons and for arbitrary polarization of all other particles is\ncalculated. We obtain the complete set of functions which describe such a\nprobability in a compact invariant form. Besides, we discuss in some detail the\npolarization effects in the kinematics relevant to the problem of electron to\nphoton conversion at photon-photon and electron-photon colliders.\n" ]

[ " A new approach for the approximation of the channel log-likelihood ratio\n(LLR) for wireless channels based on Taylor series is proposed. The\napproximation is applied to the uncorrelated flat Rayleigh fading channel with\nunknown channel state information at the receiver. It is shown that the\nproposed approximation greatly simplifies the calculation of channel LLRs, and\nyet provides results almost identical to those based on the exact calculation\nof channel LLRs. The results are obtained in the context of bit-interleaved\ncoded modulation (BICM) schemes with low-density parity-check (LDPC) codes, and\ninclude threshold calculations and error rate performance of finite-length\ncodes. Compared to the existing approximations, the proposed method is either\nsignificantly less complex, or considerably more accurate.\n", " Semantic communication allows the receiver to know the intention instead of\nthe bit information itself, which is an emerging technique to support real-time\nhuman-machine and machine-to-machine interactions for future wireless\ncommunications. In semantic communications, both transmitter and receiver share\nsome common knowledge, which can be used to extract small-size information at\nthe transmitter and recover the original information at the receiver. Due to\ndifferent design purposes, security issues in semantic communications have two\nunique features compared to standard bit-wise communications. First, an\nattacker in semantic communications considers not only the amount of stolen\ndata but also the meanings of stolen data. Second, an attacker in semantic\ncommunication systems can attack not only semantic information transmission as\ndone in standard communication systems but also attacks machine learning (ML)\nmodels used for semantic information extraction since most of semantic\ninformation is generated using ML based methods. Due to these unique features,\nin this paper, we present an overview on the fundamentals and key challenges in\nthe design of secure semantic communication. We first provide various methods\nto define and extract semantic information. Then, we focus on secure semantic\ncommunication techniques in two areas: information security and semantic ML\nmodel security. For each area, we identify the main problems and challenges.\nThen, we will provide a comprehensive treatment of these problems. In a\nnutshell,this article provides a holistic set of guidelines on how to design\nsecure semantic communication systems over real-world wireless communication\nnetworks.\n", " Recent interest in integrated sensing and communications has led to the\ndesign of novel signal processing techniques to recover information from an\noverlaid radar-communications signal. Here, we focus on a spectral coexistence\nscenario, wherein the channels and transmit signals of both radar and\ncommunications systems are unknown to the common receiver. In this dual-blind\ndeconvolution (DBD) problem, the receiver admits a multi-carrier wireless\ncommunications signal that is overlaid with the radar signal reflected off\nmultiple targets. The communications and radar channels are represented by\ncontinuous-valued range-times or delays corresponding to multiple transmission\npaths and targets, respectively. Prior works addressed recovery of unknown\nchannels and signals in this ill-posed DBD problem through atomic norm\nminimization but contingent on individual minimum separation conditions for\nradar and communications channels. In this paper, we provide an optimal joint\nseparation condition using extremal functions from the Beurling-Selberg\ninterpolation theory. Thereafter, we formulate DBD as a low-rank modified\nHankel matrix retrieval and solve it via nuclear norm minimization. We estimate\nthe unknown target and communications parameters from the recovered low-rank\nmatrix using multiple signal classification (MUSIC) method. We show that the\njoint separation condition also guarantees that the underlying Vandermonde\nmatrix for MUSIC is well-conditioned. Numerical experiments validate our\ntheoretical findings.\n", " In this study, a hybrid method based on an Adaptive Neuro-Fuzzy Inference\nSystem (ANFIS) and Particle Swarm Optimization (PSO) for diagnosing Liver\ndisorders (ANFIS-PSO) is introduced. This smart diagnosis method deals with a\ncombination of making an inference system and optimization process which tries\nto tune the hyper-parameters of ANFIS based on the data-set. The Liver diseases\ncharacteristics are taken from the UCI Repository of Machine Learning\nDatabases. The number of these characteristic attributes are 7, and the sample\nnumber is 354. The right diagnosis performance of the ANFIS-PSO intelligent\nmedical system for liver disease is evaluated by using classification accuracy,\nsensitivity and specificity analysis, respectively. According to the\nexperimental results, the performance of ANFIS-PSO can be more considerable\nthan traditional FIS and ANFIS without optimization phase.\n", " The gains afforded by cloud radio access network (C-RAN) in terms of savings\nin capital and operating expenses, flexibility, interference management and\nnetwork densification rely on the presence of high-capacity low-latency\nfronthaul connectivity between remote radio heads (RRHs) and baseband unit\n(BBU). In light of the non-uniform and limited availability of fiber optics\ncables, the bandwidth constraints on the fronthaul network call, on the one\nhand, for the development of advanced baseband compression strategies and, on\nthe other hand, for a closer investigation of the optimal functional split\nbetween RRHs and BBU. In this chapter, after a brief introduction to signal\nprocessing challenges in C-RAN, this optimal function split is studied at the\nphysical (PHY) layer as it pertains to two key baseband signal processing\nsteps, namely channel estimation in the uplink and channel encoding/ linear\nprecoding in the downlink. Joint optimization of baseband fronthaul compression\nand of baseband signal processing is tackled under different PHY functional\nsplits, whereby uplink channel estimation and downlink channel encoding/ linear\nprecoding are carried out either at the RRHs or at the BBU. The analysis, based\non information-theoretical arguments, and numerical results yields insight into\nthe configurations of network architecture and fronthaul capacities in which\ndifferent functional splits are advantageous. The treatment also emphasizes the\nversatility of deterministic and stochastic successive convex approximation\nstrategies for the optimization of C-RANs.\n", " Stochastic gradient descent (SGD) is a powerful method for large-scale\noptimization problems in the area of machine learning, especially for a\nfinite-sum formulation with numerous variables. In recent years, mini-batch SGD\ngains great success and has become a standard technique for training deep\nneural networks fed with big amount of data. Inspired by its success in deep\nlearning, we apply the idea of SGD with batch selection of samples to a classic\noptimization problem in decision version. Given $n$ unit circles, the equal\ncircle packing problem (ECPP) asks whether there exist a feasible packing that\ncould put all the circles inside a circular container without overlapping.\nSpecifically, we propose a stochastic item descent method (SIDM) for ECPP in\nlarge scale, which randomly divides the unit circles into batches and runs\nBroyden-Fletcher-Goldfarb-Shanno (BFGS) algorithm on the corresponding batch\nfunction iteratively to speedup the calculation. We also increase the batch\nsize during the batch iterations to gain higher quality solution. Comparing to\nthe current best packing algorithms, SIDM greatly speeds up the calculation of\noptimization process and guarantees the solution quality for large scale\ninstances with up to 1500 circle items, while the baseline algorithms usually\nhandle about 300 circle items. The results indicate the highly efficiency of\nSIDM for this classic optimization problem in large scale, and show potential\nfor other large scale classic optimization problems in which gradient descent\nis used for optimization.\n", " A candidate explanation of the good empirical performance of deep neural\nnetworks is the implicit regularization effect of first order optimization\nmethods. Inspired by this, we prove a convergence theorem for nonconvex\ncomposite optimization, and apply it to a general learning problem covering\nmany machine learning applications, including supervised learning. We then\npresent a deep multilayer perceptron model and prove that, when sufficiently\nwide, it $(i)$ leads to the convergence of gradient descent to a global optimum\nwith a linear rate, $(ii)$ benefits from the implicit regularization effect of\ngradient descent, $(iii)$ is subject to novel bounds on the generalization\nerror, $(iv)$ exhibits the lazy training phenomenon and $(v)$ enjoys learning\nrate transfer across different widths. The corresponding coefficients, such as\nthe convergence rate, improve as width is further increased, and depend on the\neven order moments of the data generating distribution up to an order depending\non the number of layers. The only non-mild assumption we make is the\nconcentration of the smallest eigenvalue of the neural tangent kernel at\ninitialization away from zero, which has been shown to hold for a number of\nless general models in contemporary works. We present empirical evidence\nsupporting this assumption as well as our theoretical claims.\n", " We consider the problem of assembling a sequence based on a collection of its\nsubstrings observed through a noisy channel. The mathematical basis of the\nproblem is the construction and design of sequences that may be discriminated\nbased on a collection of their substrings observed through a noisy channel. We\nexplain the connection between the sequence reconstruction problem and the\nproblem of DNA synthesis and sequencing, and introduce the notion of a DNA\nstorage channel. We analyze the number of sequence equivalence classes under\nthe channel mapping and propose new asymmetric coding techniques to combat the\neffects of synthesis and sequencing noise. In our analysis, we make use of\nrestricted de Bruijn graphs and Ehrhart theory for rational polytopes.\n", " In this paper, we tested several sparse optimization algorithms based on the\npublic dataset of the DREAM5 Gene Regulatory Network Inference Challenge. And\nwe find that introducing 20% of the regulatory network as a priori known data\ncan provide a basis for parameter selection of inference algorithms, thus\nimproving prediction efficiency and accuracy. In addition to testing common\nsparse optimization methods, we also developed voting algorithms by bagging\nthem. Experiments on the DREAM5 dataset show that the sparse optimization-based\ninference of the moderation relation works well, achieving better results than\nthe official DREAM5 results on three datasets. However, the performance of\ntraditional independent algorithms varies greatly in the face of different\ndatasets, while our voting algorithm achieves the best results on three of the\nfour datasets.\n", " Consider the following distributed optimization scenario. A worker has access\nto training data that it uses to compute the gradients while a server decides\nwhen to stop iterative computation based on its target accuracy or delay\nconstraints. The server receives all its information about the problem instance\nfrom the worker via a rate-limited noiseless communication channel. We\nintroduce the principle we call Differential Quantization (DQ) that prescribes\ncompensating the past quantization errors to direct the descent trajectory of a\nquantized algorithm towards that of its unquantized counterpart. Assuming that\nthe objective function is smooth and strongly convex, we prove that\nDifferentially Quantized Gradient Descent (DQ-GD) attains a linear contraction\nfactor of $\\max\\{\\sigma_{\\mathrm{GD}}, \\rho_n 2^{-R}\\}$, where\n$\\sigma_{\\mathrm{GD}}$ is the contraction factor of unquantized gradient\ndescent (GD), $\\rho_n \\geq 1$ is the covering efficiency of the quantizer, and\n$R$ is the bitrate per problem dimension $n$. Thus at any $R\\geq\\log_2 \\rho_n\n/\\sigma_{\\mathrm{GD}}$ bits, the contraction factor of DQ-GD is the same as\nthat of unquantized GD, i.e., there is no loss due to quantization. We show\nthat no algorithm within a certain class can converge faster than\n$\\max\\{\\sigma_{\\mathrm{GD}}, 2^{-R}\\}$. Since quantizers exist with $\\rho_n \\to\n1$ as $n \\to \\infty$ (Rogers, 1963), this means that DQ-GD is asymptotically\noptimal. The principle of differential quantization continues to apply to\ngradient methods with momentum such as Nesterov's accelerated gradient descent,\nand Polyak's heavy ball method. For these algorithms as well, if the rate is\nabove a certain threshold, there is no loss in contraction factor obtained by\nthe differentially quantized algorithm compared to its unquantized counterpart.\nExperimental results on least-squares problems validate our theoretical\nanalysis.\n", " Beyond 5G and 6G networks are expected to support new and challenging use\ncases and applications that depend on a certain level of Quality of Service\n(QoS) to operate smoothly. Predicting the QoS in a timely manner is of high\nimportance, especially for safety-critical applications as in the case of\nvehicular communications. Although until recent years the QoS prediction has\nbeen carried out by centralized Artificial Intelligence (AI) solutions, a\nnumber of privacy, computational, and operational concerns have emerged.\nAlternative solutions have been surfaced (e.g. Split Learning, Federated\nLearning), distributing AI tasks of reduced complexity across nodes, while\npreserving the privacy of the data. However, new challenges rise when it comes\nto scalable distributed learning approaches, taking into account the\nheterogeneous nature of future wireless networks. The current work proposes\nDISTINQT, a privacy-aware distributed learning framework for QoS prediction.\nOur framework supports multiple heterogeneous nodes, in terms of data types and\nmodel architectures, by sharing computations across them. This, enables the\nincorporation of diverse knowledge into a sole learning process that will\nenhance the robustness and generalization capabilities of the final QoS\nprediction model. DISTINQT also contributes to data privacy preservation by\nencoding any raw input data into a non-linear latent representation before any\ntransmission. Evaluation results showcase that our framework achieves a\nstatistically identical performance compared to its centralized version and an\naverage performance improvement of up to 65% against six state-of-the-art\ncentralized baseline solutions in the Tele-Operated Driving use case.\n", " We model biochemical signal transduction, based on a ligand-receptor binding\nmechanism, as a discrete-time finite-state Markov channel, which we call the\nBIND channel. We show how to obtain the capacity of this channel, for the case\nof binary output, binary channel state, and arbitrary finite input alphabets.\nWe show that the capacity-achieving input distribution is IID. Further, we show\nthat feedback does not increase the capacity of this channel. We show how the\ncapacity of the discrete-time channel approaches the capacity of Kabanov's\nPoisson channel, in the limit of short time steps and rapid ligand release.\n", " We present a general method for the solution of the renormalization group\nequations for the non-forward parton distributions on the two-loop level in the\nflavour singlet channel based on an orthogonal polynomial reconstruction. Using\nthis formalism we study the effects of the evolution on recently proposed model\ndistribution functions.\n", " In recent years, several algorithms for system identification with neural\nstate-space models have been introduced. Most of the proposed approaches are\naimed at reducing the computational complexity of the learning problem, by\nsplitting the optimization over short sub-sequences extracted from a longer\ntraining dataset. Different sequences are then processed simultaneously within\na minibatch, taking advantage of modern parallel hardware for deep learning. An\nissue arising in these methods is the need to assign an initial state for each\nof the sub-sequences, which is required to run simulations and thus to evaluate\nthe fitting loss. In this paper, we provide insights for calibration of neural\nstate-space training algorithms based on extensive experimentation and analyses\nperformed on two recognized system identification benchmarks. Particular focus\nis given to the choice and the role of the initial state estimation. We\ndemonstrate that advanced initial state estimation techniques are really\nrequired to achieve high performance on certain classes of dynamical systems,\nwhile for asymptotically stable ones basic procedures such as zero or random\ninitialization already yield competitive performance.\n", " Remarkable effectiveness of the channel or spatial attention mechanisms for\nproducing more discernible feature representation are illustrated in various\ncomputer vision tasks. However, modeling the cross-channel relationships with\nchannel dimensionality reduction may bring side effect in extracting deep\nvisual representations. In this paper, a novel efficient multi-scale attention\n(EMA) module is proposed. Focusing on retaining the information on per channel\nand decreasing the computational overhead, we reshape the partly channels into\nthe batch dimensions and group the channel dimensions into multiple\nsub-features which make the spatial semantic features well-distributed inside\neach feature group. Specifically, apart from encoding the global information to\nre-calibrate the channel-wise weight in each parallel branch, the output\nfeatures of the two parallel branches are further aggregated by a\ncross-dimension interaction for capturing pixel-level pairwise relationship. We\nconduct extensive ablation studies and experiments on image classification and\nobject detection tasks with popular benchmarks (e.g., CIFAR-100, ImageNet-1k,\nMS COCO and VisDrone2019) for evaluating its performance.\n", " Few-shot learning has become essential for producing models that generalize\nfrom few examples. In this work, we identify that metric scaling and metric\ntask conditioning are important to improve the performance of few-shot\nalgorithms. Our analysis reveals that simple metric scaling completely changes\nthe nature of few-shot algorithm parameter updates. Metric scaling provides\nimprovements up to 14% in accuracy for certain metrics on the mini-Imagenet\n5-way 5-shot classification task. We further propose a simple and effective way\nof conditioning a learner on the task sample set, resulting in learning a\ntask-dependent metric space. Moreover, we propose and empirically test a\npractical end-to-end optimization procedure based on auxiliary task co-training\nto learn a task-dependent metric space. The resulting few-shot learning model\nbased on the task-dependent scaled metric achieves state of the art on\nmini-Imagenet. We confirm these results on another few-shot dataset that we\nintroduce in this paper based on CIFAR100. Our code is publicly available at\nhttps://github.com/ElementAI/TADAM.\n", " Audio applications involving environmental sound analysis increasingly use\ngeneral-purpose audio representations, also known as embeddings, for transfer\nlearning. Recently, Holistic Evaluation of Audio Representations (HEAR)\nevaluated twenty-nine embedding models on nineteen diverse tasks. However, the\nevaluation's effectiveness depends on the variation already captured within a\ngiven dataset. Therefore, for a given data domain, it is unclear how the\nrepresentations would be affected by the variations caused by myriad\nmicrophones' range and acoustic conditions -- commonly known as channel\neffects. We aim to extend HEAR to evaluate invariance to channel effects in\nthis work. To accomplish this, we imitate channel effects by injecting\nperturbations to the audio signal and measure the shift in the new (perturbed)\nembeddings with three distance measures, making the evaluation domain-dependent\nbut not task-dependent. Combined with the downstream performance, it helps us\nmake a more informed prediction of how robust the embeddings are to the channel\neffects. We evaluate two embeddings -- YAMNet, and OpenL3 on monophonic\n(UrbanSound8K) and polyphonic (SONYC-UST) urban datasets. We show that one\ndistance measure does not suffice in such task-independent evaluation. Although\nFr\\'echet Audio Distance (FAD) correlates with the trend of the performance\ndrop in the downstream task most accurately, we show that we need to study FAD\nin conjunction with the other distances to get a clear understanding of the\noverall effect of the perturbation. In terms of the embedding performance, we\nfind OpenL3 to be more robust than YAMNet, which aligns with the HEAR\nevaluation.\n", " One of the most important parts of Artificial Neural Networks is minimizing\nthe loss functions which tells us how good or bad our model is. To minimize\nthese losses we need to tune the weights and biases. Also to calculate the\nminimum value of a function we need gradient. And to update our weights we need\ngradient descent. But there are some problems with regular gradient descent ie.\nit is quite slow and not that accurate. This article aims to give an\nintroduction to optimization strategies to gradient descent. In addition, we\nshall also discuss the architecture of these algorithms and further\noptimization of Neural Networks in general\n", " We perform experimental investigations of osmotically driven flows in\nartificial microchannels by studying the dynamics and structure of the front of\na sugar solution traveling in 200 um wide and 50-200 um deep microchannels. We\nfind that the sugar front travels with constant speed, and that this speed is\nproportional to the concentration of the sugar solution and inversely\nproportional to the depth of the channel. We propose a theoretical model,\nwhich, in the limit of low axial flow resistance, predicts that the sugar front\nindeed should travel with a constant velocity. The model also predicts an\ninverse relationship between the depth of the channel and the speed and a\nlinear relation between the sugar concentration and the speed. We thus find\ngood agreement between the experimental results and the predictions of the\nmodel. Our motivation for studying osmotically driven flows is that they are\nbelieved to be responsible for the translocation of sugar in plants through the\nphloem sieve element cells. Also, we suggest that osmotic elements can act as\nintegrated pumps with no movable parts in lab-on-a-chip systems.\n", " Neural network training is usually accomplished by solving a non-convex\noptimization problem using stochastic gradient descent. Although one optimizes\nover the networks parameters, the main loss function generally only depends on\nthe realization of the neural network, i.e. the function it computes. Studying\nthe optimization problem over the space of realizations opens up new ways to\nunderstand neural network training. In particular, usual loss functions like\nmean squared error and categorical cross entropy are convex on spaces of neural\nnetwork realizations, which themselves are non-convex. Approximation\ncapabilities of neural networks can be used to deal with the latter\nnon-convexity, which allows us to establish that for sufficiently large\nnetworks local minima of a regularized optimization problem on the realization\nspace are almost optimal. Note, however, that each realization has many\ndifferent, possibly degenerate, parametrizations. In particular, a local\nminimum in the parametrization space needs not correspond to a local minimum in\nthe realization space. To establish such a connection, inverse stability of the\nrealization map is required, meaning that proximity of realizations must imply\nproximity of corresponding parametrizations. We present pathologies which\nprevent inverse stability in general, and, for shallow networks, proceed to\nestablish a restricted space of parametrizations on which we have inverse\nstability w.r.t. to a Sobolev norm. Furthermore, we show that by optimizing\nover such restricted sets, it is still possible to learn any function which can\nbe learned by optimization over unrestricted sets.\n", " We explore efficient optimization of toolpaths based on multiple criteria for\nlarge instances of 3D printing problems. We first show that the minimum turn\ncost 3D printing problem is NP-hard, even when the region is a simple polygon.\nWe develop SFCDecomp, a space filling curve based decomposition framework to\nsolve large instances of 3D printing problems efficiently by solving these\noptimization subproblems independently. For the Buddha model, our framework\nbuilds toolpaths over a total of 799,716 nodes across 169 layers, and for the\nBunny model it builds toolpaths over 812,733 nodes across 360 layers. Building\non SFCDecomp, we develop a multicriteria optimization approach for toolpath\nplanning. We demonstrate the utility of our framework by maximizing or\nminimizing tool path edge overlap between adjacent layers, while jointly\nminimizing turn costs. Strength testing of a tensile test specimen printed with\ntool paths that maximize or minimize adjacent layer edge overlaps reveal\nsignificant differences in tensile strength between the two classes of prints.\n", " In this paper, we present a topology optimization (TO) framework to enable\nautomated design of mechanical components while ensuring the result can be\nmanufactured using multi-axis machining. Although TO improves the part's\nperformance, the as-designed model is often geometrically too complex to be\nmachined and the as-manufactured model can significantly vary due to machining\nconstraints that are not accounted for during TO. In other words, many of the\noptimized design features cannot be accessed by a machine tool without\ncolliding with the part (or fixtures). The subsequent post-processing to make\nthe part machinable with the given setup requires trial-and-error without\nguarantees on preserving the optimized performance. Our proposed approach is\nbased on the well-established accessibility analysis formulation using\nconvolutions in configuration space that is extensively used in spatial\nplanning and robotics. We define an 'inaccessibility measure field' (IMF) over\nthe design domain to identify non-manufacturable features and quantify their\ncontribution to non-manufacturability. The IMF is used to penalize the\nsensitivity field of performance objectives and constraints to prevent\nformation of inaccessible regions. Unlike existing discrete formulations, our\nIMF provides a continuous spatial field that is desirable for TO convergence.\nOur approach applies to arbitrary geometric complexity of the part, tools, and\nfixtures, and is highly parallelizable on multi-core architecture. We\ndemonstrate the effectiveness of our framework on benchmark and realistic\nexamples in 2D and 3D. We also show that it is possible to directly construct\nmanufacturing plans for the optimized designs based on the accessibility\ninformation.\n", " We develop an efficient algorithm, coordinate descent FCI (CDFCI), for the\nelectronic structure ground state calculation in the configuration interaction\nframework. CDFCI solves an unconstrained non-convex optimization problem, which\nis a reformulation of the full configuration interaction eigenvalue problem,\nvia an adaptive coordinate descent method with a deterministic compression\nstrategy. CDFCI captures and updates appreciative determinants with different\nfrequencies proportional to their importance. We show that CDFCI produces\naccurate variational energy for both static and dynamic correlation by\nbenchmarking the binding curve of nitrogen dimer in the cc-pVDZ basis with\n$10^{-3}$ mHa accuracy. We also demonstrate the efficiency and accuracy of\nCDFCI for strongly correlated chromium dimer in the Ahlrichs VDZ basis and\nproduces state-of-the-art variational energy.\n", " For strongly convex objectives that are smooth, the classical theory of\ngradient descent ensures linear convergence relative to the number of gradient\nevaluations. An analogous nonsmooth theory is challenging. Even when the\nobjective is smooth at every iterate, the corresponding local models are\nunstable and the number of cutting planes invoked by traditional remedies is\ndifficult to bound, leading to convergences guarantees that are sublinear\nrelative to the cumulative number of gradient evaluations. We instead propose a\nmultipoint generalization of the gradient descent iteration for local\noptimization. While designed with general objectives in mind, we are motivated\nby a ``max-of-smooth'' model that captures the subdifferential dimension at\noptimality. We prove linear convergence when the objective is itself\nmax-of-smooth, and experiments suggest a more general phenomenon.\n", " Motivated by computational aspects of persistent homology for Vietoris-Rips\nfiltrations, we generalize a result of Eliyahu Rips on the contractibility of\nVietoris-Rips complexes of geodesic spaces for a suitable parameter depending\non the hyperbolicity of the space. We consider the notion of geodesic defect to\nextend this result to general metric spaces in a way that is also compatible\nwith the filtration. We further show that for finite tree metrics the\nVietoris-Rips complexes collapse to their corresponding subforests. We relate\nour result to modern computational methods by showing that these collapses are\ninduced by the apparent pairs gradient, which is used as an algorithmic\noptimization in Ripser, explaining its particularly strong performance on\ntree-like metric data.\n", " In wireless networks, energy consumed for communication includes both the\ntransmission and the processing energy. In this paper, point-to-point\ncommunication over a fading channel with an energy harvesting transmitter is\nstudied considering jointly the energy costs of transmission and processing.\nUnder the assumption of known energy arrival and fading profiles, optimal\ntransmission policy for throughput maximization is investigated. Assuming that\nthe transmitter has sufficient amount of data in its buffer at the beginning of\nthe transmission period, the average throughput by a given deadline is\nmaximized. Furthermore, a \"directional glue pouring algorithm\" that computes\nthe optimal transmission policy is described.\n", " The Grassmann manifold G_{n,p}(L) is the set of all p-dimensional planes\n(through the origin) in the n-dimensional Euclidean space L^{n}, where L is\neither R or C. This paper considers an unequal dimensional quantization in\nwhich a source in G_{n,p}(L) is quantized through a code in G_{n,q}(L), where p\nand q are not necessarily the same. It is different from most works in\nliterature where p\\equiv q. The analysis for unequal dimensional quantization\nis based on the volume of a metric ball in G_{n,p}(L) whose center is in\nG_{n,q}(L). Our chief result is a closed-form formula for the volume of a\nmetric ball when the radius is sufficiently small. This volume formula holds\nfor Grassmann manifolds with arbitrary n, p, q and L, while previous results\npertained only to some special cases. Based on this volume formula, several\nbounds are derived for the rate distortion tradeoff assuming the quantization\nrate is sufficiently high. The lower and upper bounds on the distortion rate\nfunction are asymptotically identical, and so precisely quantify the asymptotic\nrate distortion tradeoff. We also show that random codes are asymptotically\noptimal in the sense that they achieve the minimum achievable distortion with\nprobability one as n and the code rate approach infinity linearly. Finally, we\ndiscuss some applications of the derived results to communication theory. A\ngeometric interpretation in the Grassmann manifold is developed for capacity\ncalculation of additive white Gaussian noise channel. Further, the derived\ndistortion rate function is beneficial to characterizing the effect of\nbeamforming matrix selection in multi-antenna communications.\n", " For this year's 8TeV run of the LHC we lay out different strategies to search\nfor scalar top pairs. We show results for the hadronic and for the\nsemi-leptonic channels based on hadronic top tagging. For the di-lepton channel\nwe illustrate the impact of transverse mass variables. Each of our\nsignal-to-background ratios ranges around unity for a stop mass around 400GeV.\nThe combined signal significances show that dedicated stop searches are\nbecoming sensitive over a non-negligible part of parameter space.\n", " The problem of packing equal spheres in a spherical container is a classic\nglobal optimization problem, which has attracted enormous studies in academia\nand found various applications in industry. This problem is computationally\nchallenging, and many efforts focus on small-scale instances with the number of\nspherical items less than 200 in the literature. In this work, we propose an\nefficient local search heuristic algorithm named solution space exploring and\ndescent for solving this problem, which can quantify the solution's quality to\ndetermine the number of exploring actions and quickly discover a high-quality\nsolution. Besides, we propose an adaptive neighbor object maintenance method to\nspeed up the convergence of the continuous optimization process and reduce the\ntime consumption. Computational experiments on a large number of benchmark\ninstances with $5 \\leq n \\leq 400$ spherical items show that our algorithm\nsignificantly outperforms the state-of-the-art algorithm. In particular, it\nimproves the 274 best-known results and matches the 84 best-known results out\nof the 396 well-known benchmark instances.\n", " We establish the various properties as well as diverse relations of the\nascent and descent spectra for bounded linear operators. We specially focus on\nthe theory of subspectrum. Furthermore, we construct a new concept of\nconvergence for such spectra.\n", " Graph Visualization, also known as Graph Drawing, aims to find geometric\nembeddings of graphs that optimize certain criteria. Stress is a widely used\nmetric; stress is minimized when every pair of nodes is positioned at their\nshortest path distance. However, stress optimization presents computational\nchallenges due to its inherent complexity and is usually solved using\nheuristics in practice. We introduce a scalable Graph Neural Network (GNN)\nbased Graph Drawing framework with sub-quadratic runtime that can learn to\noptimize stress. Inspired by classical stress optimization techniques and\nforce-directed layout algorithms, we create a coarsening hierarchy for the\ninput graph. Beginning at the coarsest level, we iteratively refine and\nun-coarsen the layout, until we generate an embedding for the original graph.\nTo enhance information propagation within the network, we propose a novel\npositional rewiring technique based on intermediate node positions. Our\nempirical evaluation demonstrates that the framework achieves state-of-the-art\nperformance while remaining scalable.\n", " An analysis of known experiments on direct and indirect detection of the\nelectron-type neutrino and antineutrino has been made. The analysis based on a\nnew hypothesis that the observed physical space is formed from a finite set of\nbyuons, \"one-dimensional vectorial objects\". It is shown in the article that\nthe hypothesis for existence of neutrinos advanced by Pauli on the basis of an\nanalysis of the conservation laws, is not unquestionable since the fulfillment\nof these laws may be secured by the physical space itself (physical vacuum)\nbeing the lowest energy state of a discrete oscillating system originating in\nthe course of byuon interaction. This effect is analogous to that of\nM\\\"ossbauer. The direct experiments on detecting neutrinos are explained from\nthe existence of a new information channel due to the uncertainty interval for\ncoordinate of the four-contact byuon interaction forming the interior geometry\nof elementary particles and their properties. Given are also the results of an\nexperiment on observation of cyclic variations of the $\\beta$-decay rate, which\nconfirm the existence of said new information channel.\n", " The availability of accurate localization is critical for multi-robot\nexploration strategies; noisy or inconsistent localization causes failure in\nmeeting exploration objectives. We aim to achieve high localization accuracy\nwith contemporary exploration map belief and vice versa without needing global\nlocalization information. This paper proposes a novel simultaneous exploration\nand localization (SEAL) approach, which uses Gaussian Processes (GP)-based\ninformation fusion for maximum exploration while performing communication graph\noptimization for relative localization. Both these cross-dependent objectives\nwere integrated through the Rao-Blackwellization technique. Distributed\nlinearized convex hull optimization is used to select the next-best unexplored\nregion for distributed exploration. SEAL outperformed cutting-edge methods on\nexploration and localization performance in extensive ROS-Gazebo simulations,\nillustrating the practicality of the approach in real-world applications.\n", " Online learning algorithms have become a ubiquitous tool in the machine\nlearning toolbox and are frequently used in small, resource-constraint\nenvironments. Among the most successful online learning methods are Decision\nTree (DT) ensembles. DT ensembles provide excellent performance while adapting\nto changes in the data, but they are not resource efficient. Incremental tree\nlearners keep adding new nodes to the tree but never remove old ones increasing\nthe memory consumption over time. Gradient-based tree learning, on the other\nhand, requires the computation of gradients over the entire tree which is\ncostly for even moderately sized trees. In this paper, we propose a novel\nmemory-efficient online classification ensemble called shrub ensembles for\nresource-constraint systems. Our algorithm trains small to medium-sized\ndecision trees on small windows and uses stochastic proximal gradient descent\nto learn the ensemble weights of these `shrubs'. We provide a theoretical\nanalysis of our algorithm and include an extensive discussion on the behavior\nof our approach in the online setting. In a series of 2~959 experiments on 12\ndifferent datasets, we compare our method against 8 state-of-the-art methods.\nOur Shrub Ensembles retain an excellent performance even when only little\nmemory is available. We show that SE offers a better accuracy-memory trade-off\nin 7 of 12 cases, while having a statistically significant better performance\nthan most other methods. Our implementation is available under\nhttps://github.com/sbuschjaeger/se-online .\n", " Deep Learning and transfer learning models are being used to generate time\nseries forecasts; however, there is scarce evidence about their performance\nprediction that it is more evident for monthly time series. The purpose of this\npaper is to compare Deep Learning models with transfer learning and without\ntransfer learning and other traditional methods used for monthly forecasts to\nanswer three questions about the suitability of Deep Learning and Transfer\nLearning to generate predictions of time series. Time series of M4 and M3\ncompetitions were used for the experiments. The results suggest that deep\nlearning models based on TCN, LSTM, and CNN with transfer learning tend to\nsurpass the performance prediction of other traditional methods. On the other\nhand, TCN and LSTM, trained directly on the target time series, got similar or\nbetter performance than traditional methods for some forecast horizons.\n", " We introduce a notion of {\\em cyclic Schur-positivity} for sets of\npermutations, which naturally extends the classical notion of Schur-positivity,\nand it involves the existence of a bijection from permutations to standard\nYoung tableaux that preserves the cyclic descent set. Cyclic Schur-positive\nsets of permutations are always Schur-positive, but the converse does not hold,\nas exemplified by inverse descent classes, Knuth classes and conjugacy classes.\n In this paper we show that certain classes of permutations invariant under\neither horizontal or vertical rotation are cyclic Schur-positive. The proof\nunveils a new equidistribution phenomenon of descent sets on permutations,\nprovides affirmative solutions to conjectures from [9] and [2], and yields new\nexamples of Schur-positive sets.\n", " Real-valued functions on geometric data -- such as node attributes on a graph\n-- can be optimized using descriptors from persistent homology, allowing the\nuser to incorporate topological terms in the loss function. When optimizing a\nsingle real-valued function (the one-parameter setting), there is a canonical\nchoice of descriptor for persistent homology: the barcode. The operation\nmapping a real-valued function to its barcode is differentiable almost\neverywhere, and the convergence of gradient descent for losses using barcodes\nis relatively well understood. When optimizing a vector-valued function (the\nmultiparameter setting), there is no unique choice of descriptor for\nmultiparameter persistent homology, and many distinct descriptors have been\nproposed. This calls for the development of a general framework for\ndifferentiability and optimization that applies to a wide range of\nmultiparameter homological descriptors. In this article, we develop such a\nframework and show that it encompasses well-known descriptors of different\nflavors, such as signed barcodes and the multiparameter persistence landscape.\nWe complement the theory with numerical experiments supporting the idea that\noptimizing multiparameter homological descriptors can lead to improved\nperformances compared to optimizing one-parameter descriptors, even when using\nthe simplest and most efficiently computable multiparameter descriptors.\n", " Two notions of riffle shuffling on finite Coxeter groups are given: one using\nSolomon's descent algebra and another using random walk on chambers of\nhyperplane arrangements. These coincide for types $A$,$B$,$C$, $H_3$, and rank\ntwo groups. Both notions have the same, simple eigenvalues. The hyperplane\ndefinition is especially natural and satisfies a positivity property when $W$\nis crystallographic and the relevant parameter is a good prime. The hyperplane\nviewpoint suggests interesting connections with Lie theory and leads to a\nnotion of riffle shuffling for arbitrary real hyperplane arrangements and\noriented matroids. Connections with Cellini's descent algebra are given.\n", " Understanding the skill sets and knowledge required for any career is of\nutmost importance, but it is increasingly challenging in today's dynamic world\nwith rapid changes in terms of the tools and techniques used. Thus, it is\nespecially important to be able to accurately identify the required skill sets\nfor any job for better career insights and development. In this paper, we\npropose and develop the Skill Recommendation (SkillRec) system for recommending\nthe relevant job skills required for a given job based on the job title.\nSkillRec collects and identify the skill set required for a job based on the\njob descriptions published by companies hiring for these roles. In addition to\nthe data collection and pre-processing capabilities, SkillRec also utilises\nword/sentence embedding techniques for job title representation, alongside a\nfeed-forward neural network for job skill recommendation based on the job title\nrepresentation. Based on our preliminary experiments on a dataset of 6,000 job\ntitles and descriptions, SkillRec shows a promising performance in terms of\naccuracy and F1-score.\n", " Given $n$ points in a circular region $C$ in the plane, we study the problems\nof moving the $n$ points to its boundary to form a regular $n$-gon such that\nthe maximum (min-max) or the sum (min-sum) of the Euclidean distances traveled\nby the points is minimized. The problems have applications, e.g., in mobile\nsensor barrier coverage of wireless sensor networks. The min-max problem\nfurther has two versions: the decision version and optimization version. For\nthe min-max problem, we present an $O(n\\log^2 n)$ time algorithm for the\ndecision version and an $O(n\\log^3 n)$ time algorithm for the optimization\nversion. The previously best algorithms for the two problem versions take\n$O(n^{3.5})$ time and $O(n^{3.5}\\log n)$ time, respectively. For the min-sum\nproblem, we show that a special case with all points initially lying on the\nboundary of the circular region can be solved in $O(n^2)$ time, improving a\nprevious $O(n^4)$ time solution. For the general min-sum problem, we present a\n3-approximation $O(n^2)$ time algorithm, improving the previous\n$(1+\\pi)$-approximation $O(n^2)$ time algorithm. A by-product of our techniques\nis an algorithm for dynamically maintaining the maximum matching of a circular\nconvex bipartite graph; our algorithm can handle each vertex insertion or\ndeletion on the graph in $O(\\log^2 n)$ time. This result is interesting in its\nown right.\n", " In recent years, Generative Adversarial Networks (GANs) have drawn a lot of\nattentions for learning the underlying distribution of data in various\napplications. Despite their wide applicability, training GANs is notoriously\ndifficult. This difficulty is due to the min-max nature of the resulting\noptimization problem and the lack of proper tools of solving general\n(non-convex, non-concave) min-max optimization problems. In this paper, we try\nto alleviate this problem by proposing a new generative network that relies on\nthe use of random discriminators instead of adversarial design. This design\nhelps us to avoid the min-max formulation and leads to an optimization problem\nthat is stable and could be solved efficiently. The performance of the proposed\nmethod is evaluated using handwritten digits (MNIST) and Fashion products\n(Fashion-MNIST) data sets. While the resulting images are not as sharp as\nadversarial training, the use of random discriminator leads to a much faster\nalgorithm as compared to the adversarial counterpart. This observation, at the\nminimum, illustrates the potential of the random discriminator approach for\nwarm-start in training GANs.\n", " It is a standing open question to decide whether the Yao-Yao structure for\nunit disk graphs (UDGs) is a length spanner of not. This question is highly\nrelevant to the topology control problem for wireless ad hoc networks. In this\npaper we make progress towards resolving this question by showing that the\nYao-Yao structure is a length spanner for UDGs of bounded aspect ratio. We also\npropose a new local algorithm, called Yao-Sparse-Sink, based on the Yao-Sink\nmethod introduced by Li, Wan, Wang and Frieder, that computes a (1+e)-spanner\nof bounded degree for a given UDG and for given e > 0. The Yao-Sparse-Sink\nmethod enables an efficient local computation of sparse sink trees. Finally, we\nshow that all these structures for UDGs -- Yao, Yao-Yao, Yao-Sink and\nYao-Sparse-Sink -- have arbitrarily large weight.\n", " Usage of mobile sink(s) for data gathering in wireless sensor networks(WSNs)\nimproves the performance of WSNs in many respects such as power consumption,\nlifetime, etc. In some applications, the mobile sink $MS$ travels along a\npredefined path to collect data from the nearby sensors, which are referred as\nsub-sinks. Due to the slow speed of the $MS$, the data delivery latency is\nhigh. However, optimizing the {\\em data gathering schedule}, i.e. optimizing\nthe transmission schedule of the sub-sinks to the $MS$ and the movement speed\nof the $MS$ can reduce data gathering latency. We formulate two novel\noptimization problems for data gathering in minimum time. The first problem\ndetermines an optimal data gathering schedule of the $MS$ by controlling the\ndata transmission schedule and the speed of the $MS$, where the data\navailabilities of the sub-sinks are given. The second problem generalizes the\nfirst, where the data availabilities of the sub-sinks are unknown. Plane sweep\nalgorithms are proposed for finding optimal data gathering schedule and data\navailabilities of the sub-sinks. The performances of the proposed algorithms\nare evaluated through simulations. The simulation results reveal that the\noptimal distribution of data among the sub-sinks together with optimal data\ngathering schedule improves the data gathering time.\n", " We review models and approaches for error-control in order to prevent the\nbuildup of noise when gates for digital chemical and biomolecular computing\nbased on (bio)chemical reaction processes are utilized to realize stable,\nscalable networks for information processing. Solvable rate-equation models\nillustrate several recently developed methodologies for gate-function\noptimization. We also survey future challenges and possible new research\navenues.\n", " Solar activity It is vital to accurately predict solar activity, in order to\ndecrease the plausible damage of electronic equipment in the event of a large\nhigh-intensity solar eruption. Recently, we have proposed BELFIS (Brain\nEmotional Learning-based Fuzzy Inference System) as a tool for the forecasting\nof chaotic systems. The structure of BELFIS is designed based on the neural\nstructure of fear conditioning. The function of BELFIS is implemented by\nassigning adaptive networks to the components of the BELFIS structure. This\npaper especially focuses on performance evaluation of BELFIS as a predictor by\nforecasting solar cycles 16 to 24. The performance of BELFIS is compared with\nother computational models used for this purpose, and in particular with\nadaptive neuro-fuzzy inference system (ANFIS).\n", " A result of Foata and Schutzenberger states that two statistics on\npermutations, the number of inversions and the inverse major index, have the\nsame distribution on a descent class. We give a multivariate generalization of\nthis property: the sorted vectors of the Lehmer code, of the inverse majcode,\nand of a new code (the inverse saillance code), have the same distribution on a\ndescent class, and their common multivariate generating function is a flagged\nribbon Schur function.\n", " This paper presents a new algorithm for the fast, shared memory, multi-core\ncomputation of augmented contour trees on triangulations. In contrast to most\nexisting parallel algorithms our technique computes augmented trees, enabling\nthe full extent of contour tree based applications including data segmentation.\nOur approach completely revisits the traditional, sequential contour tree\nalgorithm to re-formulate all the steps of the computation as a set of\nindependent local tasks. This includes a new computation procedure based on\nFibonacci heaps for the join and split trees, two intermediate data structures\nused to compute the contour tree, whose constructions are efficiently carried\nout concurrently thanks to the dynamic scheduling of task parallelism. We also\nintroduce a new parallel algorithm for the combination of these two trees into\nthe output global contour tree. Overall, this results in superior time\nperformance in practice, both in sequential and in parallel thanks to the\nOpenMP task runtime. We report performance numbers that compare our approach to\nreference sequential and multi-threaded implementations for the computation of\naugmented merge and contour trees. These experiments demonstrate the run-time\nefficiency of our approach and its scalability on common workstations. We\ndemonstrate the utility of our approach in data segmentation applications.\n", " The world is at the cusp of a new era where pivotal importance is being\nattached to the development of sustainable and high-performance energy storage\nsystems. Potassium-ion batteries are deemed not only as cheap battery\ncandidates, but also as the penultimate high-voltage energy storage systems\nwithin the monovalent-cation chemistries. However, their performance and\nsustainability are undermined by the lack of suitable electrolytes for\nhigh-voltage operation particularly due to the limited availability of cathode\nmaterials. Here, the potential of ionic liquids based on potassium\nbis(trifluoromethanesulfonyl)amide (KTFSA) as high-voltage electrolytes is\npresented by assessing their physicochemical properties, along with the\nelectrochemical properties upon coupling with new high-voltage layered cathode\nmaterials. These ionic liquids demonstrate a lower redox potential for\npotassium dissolution / deposition (with a wide voltage tolerance of around\n$6.0$ $\\rm V$), placing them as feasible and safe electrolytes for high-voltage\npotassium-ion battery configuration. This is proven by matching this\nelectrolyte with new high-voltage layered cathode compositions, demonstrating\nstable electrochemical performance. The present findings of electrochemically\nstable ionic liquids based on potassium bis(trifluoromethanesulfonyl)amide will\nbolster further advancement of high-performance cathode materials, whose\nperformance at high-voltage regimes were apparently restricted by the paucity\nof suitable and compatible electrolytes.\n", " The identification of key factors such as medications, diseases, and\nrelationships within electronic health records and clinical notes has a wide\nrange of applications in the clinical field. In the N2C2 2022 competitions,\nvarious tasks were presented to promote the identification of key factors in\nelectronic health records (EHRs) using the Contextualized Medication Event\nDataset (CMED). Pretrained large language models (LLMs) demonstrated\nexceptional performance in these tasks. This study aims to explore the\nutilization of LLMs, specifically ChatGPT, for data augmentation to overcome\nthe limited availability of annotated data for identifying the key factors in\nEHRs. Additionally, different pre-trained BERT models, initially trained on\nextensive datasets like Wikipedia and MIMIC, were employed to develop models\nfor identifying these key variables in EHRs through fine-tuning on augmented\ndatasets. The experimental results of two EHR analysis tasks, namely medication\nidentification and medication event classification, indicate that data\naugmentation based on ChatGPT proves beneficial in improving performance for\nboth medication identification and medication event classification.\n", " Classifier-free guidance is a key component for enhancing the performance of\nconditional generative models across diverse tasks. While it has previously\ndemonstrated remarkable improvements for the sample quality, it has only been\nexclusively employed for diffusion models. In this paper, we integrate\nclassifier-free guidance into Flow Matching (FM) models, an alternative\nsimulation-free approach that trains Continuous Normalizing Flows (CNFs) based\non regressing vector fields. We explore the usage of \\emph{Guided Flows} for a\nvariety of downstream applications. We show that Guided Flows significantly\nimproves the sample quality in conditional image generation and zero-shot\ntext-to-speech synthesis, boasting state-of-the-art performance. Notably, we\nare the first to apply flow models for plan generation in the offline\nreinforcement learning setting, showcasing a 10x speedup in computation\ncompared to diffusion models while maintaining comparable performance.\n" ]

[ " We study the occurrence of a strong first-order electroweak phase transition\nin composite Higgs models. Minimal constructions realising this scenario are\nbased on the coset SO(6)/SO(5) which delivers an extended Higgs sector with an\nadditional scalar. In such models, a two-step phase transition can be obtained\nwith the scalar singlet acquiring a vacuum expectation value at intermediate\ntemperatures. A bonus of the Nambu-Goldstone boson nature of the scalar-sector\ndynamics is the presence of non-renormalisable Higgs interactions that can\ntrigger additional sources of CP violation needed to realise baryogenesis at\nthe electroweak scale. Another interesting aspect of this scenario is the\ngeneration of gravitational wave signatures that can be observed at future\nspace-based interferometers.\n", " Negative-index refraction is achieved in a lamellar composite with\nepsilon-negative (ENG) and mu-negative (MNG) materials stacked alternatively.\nBased on the effective medium approximation, simultaneously negative effective\npermittivity and permeability of such a lamellar composite are obtained\ntheoretically and further proven by full-wave simulations. Consequently, the\nfamous left-handed metamaterial comprising split ring resonators and wires is\ninterpreted as an analogy of such an ENG-MNG lamellar composite. In addition,\nbeyond the effective medium approximation, the propagating field squeezed near\nthe ENG/MNG interface is demonstrated to be left-handed surface waves with\nbackward phase velocity.\n", " Oscillations play a major role in a number of biological systems, from\npredator-prey models of ecology to circadian clocks. In this paper we focus on\nthe question of whether oscillations exist within dual-site phosphorylation\nsystems. Previously, Wang and Sontag showed, using monotone systems theory,\nthat the Michaelis-Menten (MM) approximation of the distributive and sequential\ndual-site phosphorylation system lacks oscillations. However, biological\nsystems are generally not purely distributive; there is generally some\nprocessive behavior as well. Accordingly, this paper focuses on the MM\napproximation of a general sequential dual-site phosphorylation system that\ncontains both processive and distributive components, termed the composite\nsystem. Expanding on the methods of Bozeman and Morales, we preclude\noscillations in the MM approximation of the composite system. This implies the\nlack of oscillations in the MM approximations of the processive and\ndistributive systems, shown previously, as well as in the MM approximation of\nthe partially processive and partially distributive mixed-mechanism system.\n", " By using the background field method of QCD in a path integral approach, we\nderive the equation of motion for the classical chromofield and for the gluon\nin a system containing the gluon and the classical chromofield simultaneously.\nThis inhomogeneous field equation contains a current term, which is the\nexpectation value of a composite operator including linear, square and cubic\nterms of the gluon field. We also derive identities which the current should\nobey from the gauge invariance. We calculate the current at the leading order\nwhere the current induced by the gluon is opposite in sign to that induced by\nthe quark. This is just the feature of the non-Abelian gauge field theory which\nhas asymptotic freedom. Physically, the induced current can be treated as the\n'displacement' current in the polarized vacuum, and its effect is equivalent to\nredefining the field and the coupling constant.\n", " We continue a study of unconditionally secure all-or-nothing transforms\n(AONT) begun in \\cite{St}. An AONT is a bijective mapping that constructs s\noutputs from s inputs. We consider the security of t inputs, when s-t outputs\nare known. Previous work concerned the case t=1; here we consider the problem\nfor general t, focussing on the case t=2. We investigate constructions of\nbinary matrices for which the desired properties hold with the maximum\nprobability. Upper bounds on these probabilities are obtained via a quadratic\nprogramming approach, while lower bounds can be obtained from combinatorial\nconstructions based on symmetric BIBDs and cyclotomy. We also report some\nresults on exhaustive searches and random constructions for small values of s.\n", " This paper presents a systematic study of applying composite method\napproximations with locally dense basis sets (LDBS) to efficiently calculate\nNMR shielding constants in small and medium-sized molecules. The pcSseg-n\nseries of basis sets are shown to have similar accuracy to the pcS-n series\nwhen n $\\geq1$ and can slightly reduce compute costs. We identify two different\nLDBS partition schemes that perform very effectively for density functional\ncalculations. We select a large subset of the recent NS372 database containing\n290 H, C, N, and O shielding values evaluated by reference methods on 106\nmolecules to carefully assess methods of the high, medium, and low compute\ncosts to make practical recommendations. Our assessment covers conventional\nelectronic structure methods (DFT and wavefunction) with global basis\ncalculations, as well as their use in one of the satisfactory LDBS approaches,\nand a range of composite approaches, also with and without LDBS. Altogether 99\nmethods are evaluated. On this basis, we recommend different methods to reach\nthree different levels of accuracy and time requirements across the four nuclei\nconsidered.\n", " The energy sciences attempt to meet the increasing world-wide need for\nenergy, as well as sustainability goals, by cleaner sources of energy, by new\nalternative sources of energy, and by more efficient uses of available energy.\nThese goals are entirely consistent with the principles of green chemistry.\nThis chapter concerns devices for creating electricity from light and for\ncreating light from electricity. The major focus is on the photoproperties of\nruthenium and iridium complexes, which have been proven to be a rich source of\ninspiration for conceiving photoactivated devices, including organic\nphotovoltaic (OPV) cells and organic light emitting diodes (OLEDs). The chapter\nreviews important already-in-use and potential applications of ruthenium and\niridium complex-based photodevices, including the underlying mechanism behind\ntheir functioning and its investigation through computational chemistry\napproaches. It highlights the role of the information obtained from\ncomputational studies for the design of more efficient photodevices. The final\npart complements the discourse with a review of the progress on greener\nalternatives for OPVs and OLEDs.\n", " The twentieth century has thrown up exotic concepts -- dark matter, gravity\nwaves, Higgs Bosons, Magnetic Monopoles and so on. The sad truth is that even\nafter several decades, these remain elusive to observation and experiment. Some\nare now questioning these conjectures. Their verification has become a matter\nmore of hope than conviction. We will examine some alternatives in the light of\nthe above pointing out that, on the other hand, an extra neutrino, recently\npredicted by the author may have just been discovered.\n", " This paper addresses the problem of statistical inference for latent\ncontinuous-time stochastic processes, which is often intractable, particularly\nfor discrete state space processes described by Markov jump processes. To\novercome this issue, we propose a new tractable inference scheme based on an\nentropic matching framework that can be embedded into the well-known\nexpectation propagation algorithm. We demonstrate the effectiveness of our\nmethod by providing closed-form results for a simple family of approximate\ndistributions and apply it to the general class of chemical reaction networks,\nwhich are a crucial tool for modeling in systems biology. Moreover, we derive\nclosed form expressions for point estimation of the underlying parameters using\nan approximate expectation maximization procedure. We evaluate the performance\nof our method on various chemical reaction network instantiations, including a\nstochastic Lotka-Voltera example, and discuss its limitations and potential for\nfuture improvements. Our proposed approach provides a promising direction for\naddressing complex continuous-time Bayesian inference problems.\n", " Scalar composite boson masses have been computed in QCD and Technicolor\ntheories with the help of the homogeneous Bethe-Salpeter equation (BSE),\nresulting in a scalar mass that is twice the dynamically generated fermion or\ntechnifermion mass ($m_{dyn}$). We show that in the case of walking (or\nquasi-conformal) technicolor theories, where the $m_{dyn}$ behavior with the\nmomenta may be quite different from the one predicted by the standard operator\nproduct expansion, this result is incomplete and we must consider the effect of\nthe normalization condition of the BSE to determine the scalar masses. We\ncompute the composite Higgs boson mass for several groups with technifermions\nin the fundamental and higher dimensional representations and comment about the\nexperimental constraints on these theories, which indicate that models based on\nwalking theories with fermions in the fundamental representation may, within\nthe limitations of our approach, have masses quite near the actual direct\nexclusion limit.\n", " In this paper we propose two different primal-dual splitting algorithms for\nsolving inclusions involving mixtures of composite and parallel-sum type\nmonotone operators which rely on an inexact Douglas-Rachford splitting method,\nhowever applied in different underlying Hilbert spaces. Most importantly, the\nalgorithms allow to process the bounded linear operators and the set-valued\noperators occurring in the formulation of the monotone inclusion problem\nseparately at each iteration, the latter being individually accessed via their\nresolvents. The performances of the primal-dual algorithms are emphasized via\nsome numerical experiments on location and image deblurring problems.\n", " The study of knots and links from a probabilistic viewpoint provides insight\ninto the behavior of \"typical\" knots, and opens avenues for new constructions\nof knots and other topological objects with interesting properties. The\nknotting of random curves arises also in applications to the natural sciences,\nsuch as in the context of the structure of polymers. We present here several\nknown and new randomized models of knots and links. We review the main known\nresults on the knot distribution in each model. We discuss the nature of these\nmodels and the properties of the knots they produce. Of particular interest to\nus are finite type invariants of random knots, and the recently studied\nPetaluma model. We report on rigorous results and numerical experiments\nconcerning the asymptotic distribution of such knot invariants. Our approach\nraises questions of universality and classification of the various random knot\nmodels.\n", " In model-based reinforcement learning (MBRL), Wan et al. (2019) showed\nconditions under which the environment model could produce the expectation of\nthe next feature vector rather than the full distribution, or a sample thereof,\nwith no loss in planning performance. Such expectation models are of interest\nwhen the environment is stochastic and non-stationary, and the model is\napproximate, such as when it is learned using function approximation. In these\ncases a full distribution model may be impractical and a sample model may be\neither more expensive computationally or of high variance. Wan et al.\nconsidered only planning for prediction to evaluate a fixed policy. In this\npaper, we treat the control case - planning to improve and find a good\napproximate policy. We prove that planning with an expectation model must\nupdate a state-value function, not an action-value function as previously\nsuggested (e.g., Sorg & Singh, 2010). This opens the question of how planning\ninfluences action selections. We consider three strategies for this and present\ngeneral MBRL algorithms for each. We identify the strengths and weaknesses of\nthese algorithms in computational experiments. Our algorithms and experiments\nare the first to treat MBRL with expectation models in a general setting.\n", " We study random composite structures considered up to symmetry that are\nsampled according to weights on the inner and outer structures. This model may\nbe viewed as an unlabelled version of Gibbs partitions and encompasses\nmultisets of weighted combinatorial objects. We describe a general setting\ncharacterized by the formation of a giant component. The collection of small\nfragments is shown toconverge in total variation toward a limit object\nfollowing a P\\'olya-Boltzmann distribution.\n", " We provide a pedagogical introduction to the two main variants of real-space\nquantum Monte Carlo methods for electronic-structure calculations: variational\nMonte Carlo (VMC) and diffusion Monte Carlo (DMC). Assuming no prior knowledge\non the subject, we review in depth the Metropolis-Hastings algorithm used in\nVMC for sampling the square of an approximate wave function, discussing details\nimportant for applications to electronic systems. We also review in detail the\nmore sophisticated DMC algorithm within the fixed-node approximation,\nintroduced to avoid the infamous Fermionic sign problem, which allows one to\nsample a more accurate approximation to the ground-state wave function.\nThroughout this review, we discuss the statistical methods used for evaluating\nexpectation values and statistical uncertainties. In particular, we show how to\nestimate nonlinear functions of expectation values and their statistical\nuncertainties.\n", " In this paper, the quantum theory of the infinite-component Majorana field\nfor the fermionic tower is formulated. This study proves that the energy states\nwith increasing spin are simply composite systems made by a bradyon and\nantitachyons with half-integer spin. The quantum field describing these exotic\nstates is obtained by the infinite sum of four-spinor operators, which each\noperator depends on the spin and the rest mass of the bradyon in its\nfundamental state. The interaction between bradyon-tachyon, tachyon-tachyon and\ntachyon-luxon has also been considered and included in the total Lagrangian.\nThe obtained theory is consistent with the CPT invariance and the\nspin-statistics theorem and could explain the existence of new forms of matter\nnot predictable within the standard model.\n", " Persistent homology is a vital tool for topological data analysis. Previous\nwork has developed some statistical estimators for characteristics of\ncollections of persistence diagrams. However, tools that provide statistical\ninference for observations that are persistence diagrams are limited.\nSpecifically, there is a need for tests that can assess the strength of\nevidence against a claim that two samples arise from the same population or\nprocess. We propose the use of randomization-style null hypothesis significance\ntests (NHST) for these situations. The test is based on a loss function that\ncomprises pairwise distances between the elements of each sample and all the\nelements in the other sample. We use this method to analyze a range of\nsimulated and experimental data. Through these examples we experimentally\nexplore the power of the p-values. Our results show that the\nrandomization-style NHST based on pairwise distances can distinguish between\nsamples from different processes, which suggests that its use for hypothesis\ntests upon persistence diagrams is reasonable. We demonstrate its application\non a real dataset of fMRI data of patients with ADHD.\n", " A Condorcet domain (CD) is a collection of linear orders on a set of\ncandidates satisfying the following property: for any choice of preferences of\nvoters from this collection, a simple majority rule does not yield cycles. We\npropose a method of constructing \"large\" CDs by use of rhombus tiling diagrams\nand explain that this method unifies several constructions of CDs known\nearlier. Finally, we show that three conjectures on the maximal sizes of those\nCDs are, in fact, equivalent and provide a counterexample to them.\n", " Certain scalar-tensor theories of gravity provide negative-energy, tachyonic\nmodes to a fundamental scalar inside matter, giving rise to non-perturbative\nphenomena around compact stars. Studies of this and other tachyonic\ninstabilities always average over local matter properties. We use elementary,\nflat space models to understand possible collective effects and the accuracy of\nthe averaging procedure. In particular, we consider bodies made of elementary\nconstituents which do not, in isolation, scalarize because their compactness\n${\\cal C}$ is too small, ${\\cal C}\\lesssim {\\cal C}_{\\rm crit}$. We show that\nwhen the individual constituents have compactness smaller but close to the\nthreshold, one is able to scalarize composite bodies through collective\neffects, and the compactness of the composite body can be made arbitrarily\nsmall. On the other hand, our results suggest that when the fundamental\nbuilding blocks have very low compactness, then scalarization of the composite\nbody requires a global compactness ${\\cal C}_{\\rm global}\\gtrsim {\\cal C}_{\\rm\ncrit}$. Thus, our results rule out scalarization of dilute bodies via\ncollective effects.\n", " In this paper, we explore different ways of training a model for handwritten\ntext recognition when multiple imperfect or noisy transcriptions are available.\nWe consider various training configurations, such as selecting a single\ntranscription, retaining all transcriptions, or computing an aggregated\ntranscription from all available annotations. In addition, we evaluate the\nimpact of quality-based data selection, where samples with low agreement are\nremoved from the training set. Our experiments are carried out on municipal\nregisters of the city of Belfort (France) written between 1790 and 1946. %\nresults The results show that computing a consensus transcription or training\non multiple transcriptions are good alternatives. However, selecting training\nsamples based on the degree of agreement between annotators introduces a bias\nin the training data and does not improve the results. Our dataset is publicly\navailable on Zenodo: https://zenodo.org/record/8041668.\n", " Discrepancies and accords between quantum (QM) and classical mechanics (CM)\nrelated to expectation values and periods are found for both the simple\nharmonic oscillator (SHO) and a free particle in a box (FPB), which may apply\ngenerally. These indicate non-locality is expected throughout QM. The FPB\nenergy states violate the Correspondence Principle. Previously unexpected\naccords are found and proven that the classical and quantum expectation values\nare the same for the expectation value of the second moment and the beat period\n(i.e. beats between the phases for adjoining energy states) for the SHO for all\nquantum numbers, n. However, for the FPB the beat periods differ significantly\nat small n. It is shown that a particle's velocity in an infinite square well\nvaries, no matter how wide the box, nor how far the particle is from the walls.\nThe quantum free particle variances share an indirect commonality with the\nAharonov-Bohm and Aharonov-Casher effects in that there is a quantum action in\nthe absence of a force. The concept of an \"Expectation Value over a Partial\nWell Width\" is introduced. This paper raises the question as to whether these\ninconsistencies are undetectable, or can be empirically ascertained. These\ninherent variances may need to be fixed, or nature is manifestly more\nnon-classical than expected.\n", " In this paper, we demonstrate how deterministic and stochastic dynamics on\nmanifolds, as well as differential geometric constructions can be implemented\nconcisely and efficiently using modern computational frameworks that mix\nsymbolic expressions with efficient numerical computations. In particular, we\nuse the symbolic expression and automatic differentiation features of the\npython library Theano, originally developed for high-performance computations\nin deep learning. We show how various aspects of differential geometry and Lie\ngroup theory, connections, metrics, curvature, left/right invariance, geodesics\nand parallel transport can be formulated with Theano using the automatic\ncomputation of derivatives of any order. We will also show how symbolic\nstochastic integrators and concepts from non-linear statistics can be\nformulated and optimized with only a few lines of code. We will then give\nexplicit examples on low-dimensional classical manifolds for visualization and\ndemonstrate how this approach allows both a concise implementation and\nefficient scaling to high dimensional problems.\n", " Gravitational-wave experiments with interferometers and with resonant masses\ncan search for stochastic backgrounds of gravitational waves of cosmological\norigin. We review both experimental and theoretical aspects of the search for\nthese backgrounds. We give a pedagogical derivation of the various relations\nthat characterize the response of a detector to a stochastic background. We\ndiscuss the sensitivities of the large interferometers under constructions\n(LIGO, VIRGO, GEO600, TAMA300, AIGO) or planned (Avdanced LIGO, LISA) and of\nthe presently operating resonant bars, and we give the sensitivities for\nvarious two-detectors correlations. We examine the existing limits on the\nenergy density in gravitational waves from nucleosynthesis, COBE and pulsars,\nand their effects on theoretical predictions. We discuss general theoretical\nprinciples for order-of-magnitude estimates of cosmological production\nmechanisms, and then we turn to specific theoretical predictions from\ninflation, string cosmology, phase transitions, cosmic strings and other\nmechanisms. We finally compare with the stochastic backgrounds of astrophysical\norigin.\n", " We prove a Fundamental Theorem of Finite Semidistributive Lattices (FTFSDL),\nmodelled on Birkhoff's Fundamental Theorem of Finite Distributive Lattices. Our\nFTFSDL is of the form \"A poset L is a finite semidistributive lattice if and\nonly if there exists a set Sha with some additional structure, such that L is\nisomorphic to the admissible subsets of Sha ordered by inclusion; in this case,\nSha and its additional structure are uniquely determined by L.\" The additional\nstructure on Sha is a combinatorial abstraction of the notion of torsion pairs\nfrom representation theory and has geometric meaning in the case of posets of\nregions of hyperplane arrangements. We show how the FTFSDL clarifies many\nconstructions in lattice theory, such as canonical join representations and\npassing to quotients, and how the semidistributive property interacts with\nother major classes of lattices. Many of our results also apply to infinite\nlattices.\n", " A novel method for sinogram denoise based on Generative Adversarial Networks\n(GANs) in the field of SPECT imaging is presented. Projection data from\nsoftware phantoms were used to train the proposed model. For evaluation of the\nefficacy of the method Shepp Logan based phantom, with various noise levels\nadded where used. The resulting denoised sinograms are reconstructed using\nOrdered Subset Expectation Maximization (OSEM) and compared to the\nreconstructions of the original noised sinograms. As the results show, the\nproposed method significantly denoise the sinograms and significantly improves\nthe reconstructions. Finally, to demonstrate the efficacy and capability of the\nproposed method results from real-world DAT-SPECT sinograms are presented.\n", " The most general Higgs potential of the two-Higgs-doublet model (2HDM)\ncontains three squared-mass parameters and seven quartic self-coupling\nparameters. Among these, one squared-mass parameter and three quartic coupling\nparameters are potentially complex. The Higgs potential explicitly violates CP\nsymmetry if and only if no choice of basis exists in the two-dimensional Higgs\n``flavor'' space in which all the Higgs potential parameters are real. We\nexhibit four independent potentially complex invariant (basis-independent)\ncombinations of mass and coupling parameters and show that the reality of all\nfour invariants provides the necessary and sufficient conditions for an\nexplicitly CP-conserving 2HDM scalar potential. Additional potentially complex\ninvariants can be constructed that depend on the Higgs field vacuum expectation\nvalues (vevs). We demonstrate how these can be used together with the\nvev-independent invariants to distinguish between explicit and spontaneous\nCP-violation in the Higgs sector.\n", " Gene Set Enrichment Analysis (GSEA) and its variations aim to discover\ncollections of genes that show moderate but coordinated differences in\nexpression. However, such techniques may be ineffective if many individual\ngenes in a phenotype-related gene set have weak discriminative power. A\npotential solution is to search for combinations of genes that are highly\ndifferentiating even when individual genes are not. Although such techniques\nhave been developed, these approaches have not been used with GSEA to any\nsignificant degree because of the large number of potential gene combinations\nand the heterogeneity of measures that assess the differentiation provided by\ngene groups of different sizes.\n To integrate the search for differentiating gene combinations and GSEA, we\npropose a general framework with two key components: (A) a procedure that\nreduces the number of scores to be handled by GSEA to the number of genes by\nsummarizing the scores of the gene combinations involving a particular gene in\na single score, and (B) a procedure to integrate the heterogeneous scores from\ncombinations of different sizes and from different gene combination measures by\nmapping the scores to p-values. Experiments on four gene expression data sets\ndemonstrate that the integration of GSEA and gene combination search can\nenhance the power of traditional GSEA by discovering gene sets that include\ngenes with weak individual differentiation but strong joint discriminative\npower. Also, gene sets discovered by the integrative framework share several\ncommon biological processes and improve the consistency of the results among\nthree lung cancer data sets.\n", " The study of (minimally) rigid graphs is motivated by numerous applications,\nmostly in robotics and bioinformatics. A major open problem concerns the number\nof embeddings of such graphs, up to rigid motions, in Euclidean space. We\ncapture embeddability by polynomial systems with suitable structure, so that\ntheir mixed volume, which bounds the number of common roots, to yield\ninteresting upper bounds on the number of embeddings. We focus on $\\RR^2$ and\n$\\RR^3$, where Laman graphs and 1-skeleta of convex simplicial polyhedra,\nrespectively, admit inductive Henneberg constructions. We establish the first\nlower bound in $\\RR^3$ of about $2.52^n$, where $n$ denotes the number of\nvertices. Moreover, our implementation yields upper bounds for $n \\le 10$ in\n$\\RR^2$ and $\\RR^3$, which reduce the existing gaps, and tight bounds up to\n$n=7$ in $\\RR^3$.\n", " In this paper we discuss in detail the interface between Classical\nElectrodynamics and Quantum Theory, which shows up as well known unphysical\nphenomena at the Compton scale in both the theories and argue that the photon\nof the electromagnetic field can be considered to be a composite of a massless\nDirac particle and its anti particle.\n", " Stein kernels are a way of comparing probability distributions, defined via\nintegration by parts formulas. We provide two constructions of Stein kernels in\nfree probability. One is given by an explicit formula, and the other via free\nPoincar\\'e inequalities. In particular, we show that unlike in the classical\nsetting, free Stein kernels always exist. As corollaries, we derive new bounds\non the rate of convergence in the free CLT, and a strengthening of a\ncharacterization of the semicircular law due to Biane.\n", " Previous constructions of non-type (D) maximal monotone operators were based\non the non-type (D) operators introduced by Gossez, and the construction of\nsuch operators or the proof that they were non-type (D) were not\nstraightforward. The aim of this paper is to present a very simple non-type (D)\nlinear isometry.\n", " Seminal works on light spanners over the years provide spanners with optimal\nlightness in various graph classes, such as in general graphs, Euclidean\nspanners, and minor-free graphs. Three shortcomings of previous works on light\nspanners are: (1) The techniques are ad hoc per graph class, and thus can't be\napplied broadly. (2) The runtimes of these constructions are almost always\nsub-optimal, and usually far from optimal. (3) These constructions are optimal\nin the standard and crude sense, but not in a refined sense that takes into\naccount a wider range of involved parameters.\n This work aims at addressing these shortcomings by presenting a unified\nframework of light spanners in a variety of graph classes. Informally, the\nframework boils down to a transformation from sparse spanners to light\nspanners; since the state-of-the-art for sparse spanners is much more advanced\nthan that for light spanners, such a transformation is powerful. Our framework\nis developed in two papers. The current paper is the first of the two -- it\nlays the basis of the unified framework and then applies it to design fast\nconstructions with optimal lightness for several graph classes.\n Among various applications and implications of our framework, we highlight\nhere the following:\n _ In low-dimensional Euclidean spaces, we present an $O(n\\log n)$-time\nconstruction of $(1+\\epsilon)$-spanners with lightness and degree both bounded\nby constants in the algebraic computation tree (ACT). Our construction resolves\na major problem in the area of geometric spanners, which was open for three\ndecades.\n _ In general graphs, for any $k \\geq 2$, we construct a\n$(2k-1)(1+\\epsilon)$-spanner with lightness $O(n^{1/k})$ in $O(m \\alpha(m,n))$\ntime. This result for light spanners in general weighted graphs is surprising,\nas it outperforms the analog one for sparse spanners.\n", " QCD in non-integer $d=4-2\\epsilon$ space-time dimensions enjoys conformal\ninvariance at the special fine-tuned value of the coupling. Counterterms for\ncomposite operators in minimal subtraction schemes do not depend on $\\epsilon$\nby construction, and therefore the renormalization group equations for\ncomposite operators in physical (integer) dimensions inherit conformal\nsymmetry. This observation can be used to restore the complete evolution\nkernels that take into account mixing with the operators containing total\nderivatives from their eigenvalues (anomalous dimensions). Using this approach\nwe calculate the two-loop (NLO) evolution kernels for the leading twist\nflavor-singlet operators in the position space (light-ray operator)\nrepresentation. As the main result of phenomenological relevance, in this way\nwe are able to confirm the evolution equations of flavor-singlet generalized\nhadron parton distributions derived earlier by Belitsky and M\\\"uller using a\ndifferent approach.\n", " Aprotic alkali metal-oxygen batteries are widely considered to be promising\nhigh specific energy alternatives to Li-ion batteries. The growth and\ndissolution of alkali metal oxides such as Li2O2 in Li-O2 batteries and NaO2\nand KO2 in Na- and K-O2 batteries, respectively, is central to the discharge\nand charge processes in these batteries. However, crystal growth and\ndissolution of the discharge products, especially in aprotic electrolytes, is\npoorly understood. In this work, we chose the growth of NaO2 in Na-O2 batteries\nas a model system and show that there is a strong correlation between the\nelectrolyte salt concentration and the NaO2 crystal size. With a combination of\nexperiments and theory, we argue that the correlation is a direct manifestation\nof the strong cation-anion interactions leading to decreased crystal growth\nrate at high salt concentrations. Further, we propose and experimentally\ndemonstrate that cation-coordinating crown molecules are suitable\nelectrochemically stable electrolyte additives that weaken ion-pairing and\nenhance discharge capacities in metal-oxygen batteries while not negatively\naffecting their rechargeability.\n", " We argue that one has to distinguish between the Harari-Shupe model (HSM) and\nthe Harari-Shupe observation (HSO). The former - in which quarks and leptons\nare viewed as composite objects built from confined fermionic subparticles\n(`rishons') - is known to be beset with many difficulties. The latter may be\nroughly defined as this part of the HSM that really works. We recall that the\nphase-space Clifford-algebra approach leads to the HSO without any of the HSM\nproblems and discuss in some detail how this is achieved. The light which the\nphase-space-based view sheds on the HSO sets then a new direction along which\nthe connection between space and particle properties could be studied and\noffers a glimpse into weird physics that probably lurks much deeper than the\nfield-theoretical approach of the Standard Model.\n", " In this article, we introduce a new parameterized family of topological\ninvariants, taking the form of candidate decompositions, for multi-parameter\npersistence modules. We prove that our candidate decompositions are\ncontrollable approximations: when restricting to modules that can be decomposed\ninto interval summands, we establish theoretical results about the\napproximation error between our candidate decompositions and the true\nunderlying module in terms of the standard interleaving and bottleneck\ndistances. Moreover, even when the underlying module does not admit such a\ndecomposition, our candidate decompositions are nonetheless stable invariants;\nsmall perturbations in the underlying module lead to small perturbations in the\ncandidate decomposition. Then, we introduce MMA (Multipersistence Module\nApproximation): an algorithm for computing stable instances of such invariants,\nwhich is based on fibered barcodes and exact matchings, two constructions that\nstem from the theory of single-parameter persistence. By design, MMA can handle\nan arbitrary number of filtrations, and has bounded complexity and running\ntime. Finally, we present empirical evidence validating the generalization\ncapabilities and running time speed-ups of MMA on several data sets.\n", " We propose a new approach to model composition, based on reducing several\nmodels to the same level of complexity and subsequent combining them together.\nFirstly, we suggest a set of model reduction tools that can be systematically\napplied to a given model. Secondly, we suggest a notion of a minimal complexity\nmodel. This model is the simplest one that can be obtained from the original\nmodel using these tools and still able to approximate experimental data.\nThirdly, we propose a strategy for composing the reduced models together.\nConnection with the detailed model is preserved, which can be advantageous in\nsome applications. A toolbox for model reduction and composition has been\nimplemented as part of the BioUML software and tested on the example of\nintegrating two previously published models of the CD95 (APO-1/Fas) signaling\npathways. We show that the reduced models lead to the same dynamical behavior\nof observable species and the same predictions as in the precursor models. The\ncomposite model is able to recapitulate several experimental datasets which\nwere used by the authors of the original models to calibrate them separately,\nbut also has new dynamical properties.\n", " We provide a probabilistic interpretation of attention and show that the\nstandard dot-product attention in transformers is a special case of Maximum A\nPosteriori (MAP) inference. The proposed approach suggests the use of\nExpectation Maximization algorithms for online adaptation of key and value\nmodel parameters. This approach is useful for cases in which external agents,\ne.g., annotators, provide inference-time information about the correct values\nof some tokens, e.g, the semantic category of some pixels, and we need for this\nnew information to propagate to other tokens in a principled manner. We\nillustrate the approach on an interactive semantic segmentation task in which\nannotators and models collaborate online to improve annotation efficiency.\nUsing standard benchmarks, we observe that key adaptation boosts model\nperformance ($\\sim10\\%$ mIoU) in the low feedback regime and value propagation\nimproves model responsiveness in the high feedback regime. A PyTorch layer\nimplementation of our probabilistic attention model will be made publicly\navailable here: https://github.com/apple/ml-probabilistic-attention.\n", " We give complete and exact descriptions of spaces of ultradifferentiable\nfunctions that are closed under composition with either holomorphic or\nultradifferentiable functions -- which are two distinct cases. The proof works\nby considering formal power series, and stability under differentiation is not\nrequired. As an application of the power series approach we reprove regularity\nresults for solutions of ode's and pde's without employing tedious estimates\nimploying the Fa\\`a di Bruno formula for higher derivatives of composite maps.\n", " This paper presents a new algorithm for the fast, shared memory, multi-core\ncomputation of augmented contour trees on triangulations. In contrast to most\nexisting parallel algorithms our technique computes augmented trees, enabling\nthe full extent of contour tree based applications including data segmentation.\nOur approach completely revisits the traditional, sequential contour tree\nalgorithm to re-formulate all the steps of the computation as a set of\nindependent local tasks. This includes a new computation procedure based on\nFibonacci heaps for the join and split trees, two intermediate data structures\nused to compute the contour tree, whose constructions are efficiently carried\nout concurrently thanks to the dynamic scheduling of task parallelism. We also\nintroduce a new parallel algorithm for the combination of these two trees into\nthe output global contour tree. Overall, this results in superior time\nperformance in practice, both in sequential and in parallel thanks to the\nOpenMP task runtime. We report performance numbers that compare our approach to\nreference sequential and multi-threaded implementations for the computation of\naugmented merge and contour trees. These experiments demonstrate the run-time\nefficiency of our approach and its scalability on common workstations. We\ndemonstrate the utility of our approach in data segmentation applications.\n", " Here we briefly describe some topics along the lines of projective spaces and\nrelated geometric constructions connected to linear algebra, which provide\nfundamental examples in classical geometry and analysis.\n", " We prove that every set of $n$ red and $n$ blue points in the plane contains\na red and a blue point such that every circle through them encloses at least\n$n(1-\\frac{1}{\\sqrt{2}}) -o(n)$ points of the set. This is a two-colored\nversion of a problem posed by Neumann-Lara and Urrutia. We also show that every\nset $S$ of $n$ points contains two points such that every circle passing\nthrough them encloses at most $\\lfloor{\\frac{2n-3}{3}}\\rfloor$ points of $S$.\nThe proofs make use of properties of higher order Voronoi diagrams, in the\nspirit of the work of Edelsbrunner, Hasan, Seidel and Shen on this topic.\nClosely related, we also study the number of collinear edges in higher order\nVoronoi diagrams and present several constructions.\n", " A noncommutative (nc) function in $x_1,\\dots,x_g,x_1^*,\\dots,x_g$ is called\nplurisubharmonic (plush) if its nc complex Hessian takes only positive\nsemidefinite values on an nc neighborhood of 0. The main result of this paper\nshows that an nc rational function is plush if and only if it is a composite of\na convex rational function with an analytic (no $x_j^*$) rational function. The\nproof is entirely constructive. Further, a simple computable necessary and\nsufficient condition for an nc rational function to be plush is given in terms\nof its minimal realization.\n", " We define a general notion of Gel'fand-Tsetlin algebras in the framework of\ninductive limit of the family semisimple star-algebras and the notion of\ngenralized conditional expectations. The inmverse limit of such family with\nrespect to generalized expectation is bimodule over inductive limit. The most\nimportant examples conserned to inductive families of group algebras of\nclassical series of the discrete and continuous groups. I start from some\nremarks about distinguished role of I.M.Gel'fand's conceptions in whole area.\n", " Cells measure concentrations of external ligands by capturing ligand\nmolecules with cell surface receptors. The numbers of molecules captured by\ndifferent receptors co-vary because they depend on the same extrinsic ligand\nfluctuations. However, these numbers also counter-vary due to the intrinsic\nstochasticity of chemical processes because a single molecule randomly captured\nby a receptor cannot be captured by another. Such structure of receptor\ncorrelations is generally believed to lead to an increase in information about\nthe external signal compared to the case of independent receptors. We analyze a\nsolvable model of two molecular receptors and show that, contrary to this\nwidespread expectation, the correlations have a small and negative effect on\nthe information about the ligand concentration. Further, we show that\nmeasurements that average over multiple receptors are almost as informative as\nthose that track the states of every individual one.\n", " A Cayley graph on a group $G$ has a natural edge-colouring. We say that such\na graph is CCA if every automorphism of the graph that preserves this\nedge-colouring is an element of the normaliser of the regular representation of\n$G$. A group $G$ is then said to be CCA if every Cayley graph on $G$ is CCA.\n Our main result is a characterisation of non-CCA graphs on groups that are\nSylow cyclic and whose order is not divisible by four. We also provide several\nnew constructions of non-CCA graphs.\n", " A graph is called Rank-Ramsey if (i) Its clique number is small, and (ii) The\nadjacency matrix of its complement has small rank. We initiate a systematic\nstudy of such graphs. Our main motivation is that their constructions, as well\nas proofs of their non-existence, are intimately related to the famous log-rank\nconjecture from the field of communication complexity. These investigations\nalso open interesting new avenues in Ramsey theory.\n We construct two families of Rank-Ramsey graphs exhibiting polynomial\nseparation between order and complement rank. Graphs in the first family have\nbounded clique number (as low as $41$). These are subgraphs of certain strong\nproducts, whose building blocks are derived from triangle-free strongly-regular\ngraphs. Graphs in the second family are obtained by applying Boolean functions\nto Erd\\H{o}s-R\\'enyi graphs. Their clique number is logarithmic, but their\ncomplement rank is far smaller than in the first family, about\n$\\mathcal{O}(n^{2/3})$. A key component of this construction is our\nmatrix-theoretic view of lifts.\n We also consider lower bounds on the Rank-Ramsey numbers, and determine them\nin the range where the complement rank is $5$ or less. We consider connections\nbetween said numbers and other graph parameters, and find that the two best\nknown explicit constructions of triangle-free Ramsey graphs turn out to be far\nfrom Rank-Ramsey.\n", " Large classes of dark sector models feature mass scales and couplings very\ndifferent from the ones we observe in the Standard Model (SM). Moreover, in the\nfreeze-in mechanism, often employed by the dark sector models, it is also\nrequired that the dark sector cannot be populated during the reheating process\nlike the SM. This is the so called asymmetric reheating. Such disparities in\nsizes and scales often call for dynamical explanations. In this paper, we\nexplore a scenario in which slow evolving conformal field theories (CFTs) offer\nsuch an explanation. Building on the recent work on conformal freeze-in (COFI),\nwe focus on a coupling between the Standard Model Hypercharge gauge boson and\nan anti-symmetric tensor operator in the dark CFT. We present a scenario which\ndynamically realizes the asymmetric reheating and COFI production. With a\ndetailed study of dark matter production, and taking into account limits on the\ndark matter (DM) self-interaction, warm DM bound, and constraints from the\nstellar evolution, we demonstrate that the correct relic abundance can be\nobtained with reasonable choices of parameters. The model predicts the\nexistence of a dark photon as an emergent composite particle, with a small\nkinetic mixing also determined by the CFT dynamics, which correlates it with\nthe generation of the mass scale of the dark sector. At the same time, COFI\nproduction of dark matter is very different from those freeze-in mediated by\nthe dark photon. This is an example of the physics in which a realistic dark\nsector model can often be much richer and with unexpected features.\n", " We present several constructions of paths and processes with finite quadratic\nvariation along a refining sequence of partitions, extending previous\nconstructions to the non-uniform case. We study in particular the dependence of\nquadratic variation with respect to the sequence of partitions for these\nconstructions. We identify a class of paths whose quadratic variation along a\npartition sequence is invariant under {\\it coarsening}. This class is shown to\ninclude typical sample paths of Brownian motion, but also paths which are\n$\\frac{1}{2}$-H\\\"older continuous. Finally, we show how to extend these\nconstructions to higher dimensions.\n", " We prove that any compact semi-algebraic set is homeomorphic to the solution\nspace of some art gallery problem. Previous works have established similar\nuniversality theorems, but holding only up to homotopy equivalence, rather than\nhomeomorphism, and prior to this work, the existence of art galleries even for\nsimple spaces such as the M\\\"obius strip or the three-holed torus were unknown.\nOur construction relies on an elegant and versatile gadget to copy guard\npositions with minimal overhead. It is simpler than previous constructions,\nconsisting of a single rectangular room with convex slits cut out from the\nedges. We show that both the orientable and non-orientable surfaces of genus\n$n$ admit galleries with only $O(n)$ vertices.\n" ]

[ " Implicit Neural Representations (INR) or neural fields have emerged as a\npopular framework to encode multimedia signals such as images and radiance\nfields while retaining high-quality. Recently, learnable feature grids proposed\nby Instant-NGP have allowed significant speed-up in the training as well as the\nsampling of INRs by replacing a large neural network with a multi-resolution\nlook-up table of feature vectors and a much smaller neural network. However,\nthese feature grids come at the expense of large memory consumption which can\nbe a bottleneck for storage and streaming applications. In this work, we\npropose SHACIRA, a simple yet effective task-agnostic framework for compressing\nsuch feature grids with no additional post-hoc pruning/quantization stages. We\nreparameterize feature grids with quantized latent weights and apply entropy\nregularization in the latent space to achieve high levels of compression across\nvarious domains. Quantitative and qualitative results on diverse datasets\nconsisting of images, videos, and radiance fields, show that our approach\noutperforms existing INR approaches without the need for any large datasets or\ndomain-specific heuristics. Our project page is available at\nhttp://shacira.github.io .\n", " A file repository for calculations of cross sections and kinematic\ndistributions using Monte Carlo generators for high-energy collisions is\ndiscussed. The repository is used to facilitate effective preservation and\narchiving of data from theoretical calculations, as well as for comparisons\nwith experimental data. The HepSim data library is publicly accessible and\nincludes a number of Monte Carlo event samples with Standard Model predictions\nfor current and future experiments. The HepSim project includes a software\npackage to automate the process of downloading and viewing online Monte Carlo\nevent samples. A data streaming over a network for end-user analysis is\ndiscussed.\n", " Computing Continuum (CC) systems are challenged to ensure the intricate\nrequirements of each computational tier. Given the system's scale, the Service\nLevel Objectives (SLOs) which are expressed as these requirements, must be\nbroken down into smaller parts that can be decentralized. We present our\nframework for collaborative edge intelligence enabling individual edge devices\nto (1) develop a causal understanding of how to enforce their SLOs, and (2)\ntransfer knowledge to speed up the onboarding of heterogeneous devices. Through\ncollaboration, they (3) increase the scope of SLO fulfillment. We implemented\nthe framework and evaluated a use case in which a CC system is responsible for\nensuring Quality of Service (QoS) and Quality of Experience (QoE) during video\nstreaming. Our results showed that edge devices required only ten training\nrounds to ensure four SLOs; furthermore, the underlying causal structures were\nalso rationally explainable. The addition of new types of devices can be done a\nposteriori, the framework allowed them to reuse existing models, even though\nthe device type had been unknown. Finally, rebalancing the load within a device\ncluster allowed individual edge devices to recover their SLO compliance after a\nnetwork failure from 22% to 89%.\n", " We provide a direct, intersection theoretic, argument that the Jordan models\nof an operator of class C_{0}, of its restriction to an invariant subspace, and\nof its compression to the orthogonal complement, satisfy a multiplicative form\nof the Horn inequalities, where `inequality' is replaced by `divisibility'.\nWhen one of these inequalities is saturated, we show that there exists a\nsplitting of the operator into quasidirect summands which induces similar\nsplittings for the restriction of the operator to the given invariant subspace\nand its compression to the orthogonal complement. The result is true even for\noperators acting on nonseparable Hilbert spaces. For such operators the usual\nHorn inequalities are supplemented so as to apply to all the Jordan blocks in\nthe model.\n", " The popularity and widespread use of pruning and quantization is driven by\nthe severe resource constraints of deploying deep neural networks to\nenvironments with strict latency, memory and energy requirements. These\ntechniques achieve high levels of compression with negligible impact on\ntop-line metrics (top-1 and top-5 accuracy). However, overall accuracy hides\ndisproportionately high errors on a small subset of examples; we call this\nsubset Compression Identified Exemplars (CIE). We further establish that for\nCIE examples, compression amplifies existing algorithmic bias. Pruning\ndisproportionately impacts performance on underrepresented features, which\noften coincides with considerations of fairness. Given that CIE is a relatively\nsmall subset but a great contributor of error in the model, we propose its use\nas a human-in-the-loop auditing tool to surface a tractable subset of the\ndataset for further inspection or annotation by a domain expert. We provide\nqualitative and quantitative support that CIE surfaces the most challenging\nexamples in the data distribution for human-in-the-loop auditing.\n", " A new theoretical framework is proposed for the description of baryon\nlight-cone distribution amplitudes, based on the observation that their scale\ndependence to leading logarithmic accuracy is described by a completely\nintegrable model. The physical interpretation is that one is able to find a new\nquantum number that distinguishes partonic components in the nucleon with\ndifferent scale dependence.\n", " Adapting Foundation Models (FMs) for downstream tasks through Federated\nLearning (FL) emerges a promising strategy for protecting data privacy and\nvaluable FMs. Existing methods fine-tune FM by allocating sub-FM to clients in\nFL, however, leading to suboptimal performance due to insufficient tuning and\ninevitable error accumulations of gradients. In this paper, we propose\nFederated Proxy Fine-Tuning (FedPFT), a novel method enhancing FMs adaptation\nin downstream tasks through FL by two key modules. First, the sub-FM\nconstruction module employs a layer-wise compression approach, facilitating\ncomprehensive FM fine-tuning across all layers by emphasizing those crucial\nneurons. Second, the sub-FM alignment module conducts a two-step\ndistillations-layer-level and neuron-level-before and during FL fine-tuning\nrespectively, to reduce error of gradient by accurately aligning sub-FM with FM\nunder theoretical guarantees. Experimental results on seven commonly used\ndatasets (i.e., four text and three vision) demonstrate the superiority of\nFedPFT.\n", " We develop an efficient algorithm, coordinate descent FCI (CDFCI), for the\nelectronic structure ground state calculation in the configuration interaction\nframework. CDFCI solves an unconstrained non-convex optimization problem, which\nis a reformulation of the full configuration interaction eigenvalue problem,\nvia an adaptive coordinate descent method with a deterministic compression\nstrategy. CDFCI captures and updates appreciative determinants with different\nfrequencies proportional to their importance. We show that CDFCI produces\naccurate variational energy for both static and dynamic correlation by\nbenchmarking the binding curve of nitrogen dimer in the cc-pVDZ basis with\n$10^{-3}$ mHa accuracy. We also demonstrate the efficiency and accuracy of\nCDFCI for strongly correlated chromium dimer in the Ahlrichs VDZ basis and\nproduces state-of-the-art variational energy.\n", " In this paper we use considerations of non-commutative geometry to deduce a\nmodel for QCD interactions. The model also explains within the same theoretical\nframework hitherto purely phenomenological characteristics of the quarks like\ntheir fractional charge, mass,handedness and confinement.\n", " Recently there has been renewed interest in signed distance bound\nrepresentations due to their unique properties for 3D shape modelling. This is\nespecially the case for deep learning-based bounds. However, it is beneficial\nto work with polygons in most computer-graphics applications. Thus, in this\npaper we introduce and investigate an asymptotically fast method for\ntransforming signed distance bounds into polygon meshes. This is achieved by\ncombining the principles of sphere tracing (or ray marching) with traditional\npolygonization techniques, such as Marching Cubes. We provide theoretical and\nexperimental evidence that this approach is of the $O(N^2\\log N)$ computational\ncomplexity for a polygonization grid with $N^3$ cells. The algorithm is tested\non both a set of primitive shapes as well as signed distance bounds generated\nfrom point clouds by machine learning (and represented as neural networks).\nGiven its speed, implementation simplicity and portability, we argue that it\ncould prove useful during the modelling stage as well as in shape compression\nfor storage.\n The code is available here: https://github.com/nenadmarkus/gridhopping\n", " Clarksons algorithm is a two-staged randomized algorithm for solving linear\nprograms. This algorithm has been simplified and adapted to fit the framework\nof LP-type problems. In this framework we can tackle a number of non-linear\nproblems such as computing the smallest enclosing ball of a set of points in\nR^d . In 2006, it has been shown that the algorithm in its original form works\nfor violator spaces too, which are a proper general- ization of LP-type\nproblems. It was not clear, however, whether previous simplifications of the\nalgorithm carry over to the new setting. In this paper we show the following\ntheoretical results: (a) It is shown, for the first time, that Clarksons second\nstage can be simplified. (b) The previous simplifications of Clarksons first\nstage carry over to the violator space setting. (c) Furthermore, we show the\nequivalence of violator spaces and partitions of the hypercube by hypercubes.\n", " Vertical federated learning (VFL), where data features are stored in multiple\nparties distributively, is an important area in machine learning. However, the\ncommunication complexity for VFL is typically very high. In this paper, we\npropose a unified framework by constructing coresets in a distributed fashion\nfor communication-efficient VFL. We study two important learning tasks in the\nVFL setting: regularized linear regression and $k$-means clustering, and apply\nour coreset framework to both problems. We theoretically show that using\ncoresets can drastically alleviate the communication complexity, while nearly\nmaintain the solution quality. Numerical experiments are conducted to\ncorroborate our theoretical findings.\n", " The use of large transformer-based models such as BERT, GPT, and T5 has led\nto significant advancements in natural language processing. However, these\nmodels are computationally expensive, necessitating model compression\ntechniques that reduce their size and complexity while maintaining accuracy.\nThis project investigates and applies knowledge distillation for BERT model\ncompression, specifically focusing on the TinyBERT student model. We explore\nvarious techniques to improve knowledge distillation, including experimentation\nwith loss functions, transformer layer mapping methods, and tuning the weights\nof attention and representation loss and evaluate our proposed techniques on a\nselection of downstream tasks from the GLUE benchmark. The goal of this work is\nto improve the efficiency and effectiveness of knowledge distillation, enabling\nthe development of more efficient and accurate models for a range of natural\nlanguage processing tasks.\n", " In this paper we study a new metric for comparing Betti numbers functions in\nbidimensional persistent homology, based on coherent matchings, i.e. families\nof matchings that vary in a continuous way. We prove some new results about\nthis metric, including its stability. In particular, we show that the\ncomputation of this distance is strongly related to suitable filtering\nfunctions associated with lines of slope 1, so underlining the key role of\nthese lines in the study of bidimensional persistence. In order to prove these\nresults, we introduce and study the concepts of extended Pareto grid for a\nnormal filtering function as well as of transport of a matching. As a\nby-product, we obtain a theoretical framework for managing the phenomenon of\nmonodromy in 2D persistent homology.\n", " Hypergraphs capture multi-way relationships in data, and they have\nconsequently seen a number of applications in higher-order network analysis,\ncomputer vision, geometry processing, and machine learning. In this paper, we\ndevelop theoretical foundations for studying the space of hypergraphs using\ningredients from optimal transport. By enriching a hypergraph with probability\nmeasures on its nodes and hyperedges, as well as relational information\ncapturing local and global structures, we obtain a general and robust framework\nfor studying the collection of all hypergraphs. First, we introduce a\nhypergraph distance based on the co-optimal transport framework of Redko et al.\nand study its theoretical properties. Second, we formalize common methods for\ntransforming a hypergraph into a graph as maps between the space of hypergraphs\nand the space of graphs, and study their functorial properties and Lipschitz\nbounds. Finally, we demonstrate the versatility of our Hypergraph Co-Optimal\nTransport (HyperCOT) framework through various examples.\n", " We introduce a definition of algorithmic symmetry able to capture essential\naspects of geometric symmetry. We review, study and apply a method for\napproximating the algorithmic complexity (also known as Kolmogorov-Chaitin\ncomplexity) of graphs and networks based on the concept of Algorithmic\nProbability (AP). AP is a concept (and method) capable of recursively\nenumeration all properties of computable (causal) nature beyond statistical\nregularities. We explore the connections of algorithmic complexity---both\ntheoretical and numerical---with geometric properties mainly symmetry and\ntopology from an (algorithmic) information-theoretic perspective. We show that\napproximations to algorithmic complexity by lossless compression and an\nAlgorithmic Probability-based method can characterize properties of\npolyominoes, polytopes, regular and quasi-regular polyhedra as well as\npolyhedral networks, thereby demonstrating its profiling capabilities.\n", " We derive the evolution equations for a system of neutrinos interacting among\nthemselves and with a matter background, based upon the\nBogoliubov-Born-Green-Kirkwood-Yvon (BBGKY) hierarchy. This theoretical\nframework gives an (unclosed) set of first-order coupled integro-differential\nequations governing the evolution of the reduced density matrices. By employing\nthe hierarchy, we first rederive the mean-field evolution equations for the\nneutrino one-body density matrix associated with a system of neutrinos and\nantineutrinos interacting with matter and with an anisotropic neutrino\nbackground. Then, we derive extended evolution equations to determine neutrino\nflavor conversion beyond the commonly used mean-field approximation. To this\naim we include neutrino-antineutrino pairing correlations to the two-body\ndensity matrix. The inclusion of these new contributions leads to an extended\nevolution equation for the normal neutrino density and to an equation for the\nabnormal one involving the pairing mean-field. We discuss the possible impact\nof neutrino-antineutrino correlations on neutrino flavor conversion in the\nastrophysical and cosmological environments, and possibly upon the supernova\ndynamics. Our results can be easily generalized to an arbitrary number of\nneutrino families.\n", " Simultaneous translation (ST) outputs translation while receiving the source\ninputs, and hence requires a policy to determine whether to translate a target\ntoken or wait for the next source token. The major challenge of ST is that each\ntarget token can only be translated based on the current received source\ntokens, where the received source information will directly affect the\ntranslation quality. So naturally, how much source information is received for\nthe translation of the current target token is supposed to be the pivotal\nevidence for the ST policy to decide between translating and waiting. In this\npaper, we treat the translation as information transport from source to target\nand accordingly propose an Information-Transport-based Simultaneous Translation\n(ITST). ITST quantifies the transported information weight from each source\ntoken to the current target token, and then decides whether to translate the\ntarget token according to its accumulated received information. Experiments on\nboth text-to-text ST and speech-to-text ST (a.k.a., streaming speech\ntranslation) tasks show that ITST outperforms strong baselines and achieves\nstate-of-the-art performance.\n", " We propose an efficient framework using the Dynnikov coordinates for\nhomotopy-aware multi-agent path planning in the plane. We developed a method to\ngenerate multiple homotopically distinct solutions of multi-agent path planning\nproblem in the plane by combining our framework with revised prioritized\nplanning and proved its completeness in the grid world under specific\nassumptions. Experimentally, we demonstrated the scalability of our method for\nthe number of agents. We also confirmed experimentally that homotopy-aware\nplanning contributes to avoiding locally optimal solutions when searching for\nlow-cost trajectories for a swarm of agents in a continuous environment.\n", " The discovery of close-to-star gas-giant exo-planets lends support to the\nidea of Earth's origin as a Jupiter-like gas-giant and to the consequences of\nits compression, including whole-Earth decompression dynamics that gives rise,\nwithout requiring mantle convection, to the myriad measurements and\nobservations whose descriptions are attributed to plate tectonics. I propose\nhere another, unanticipated consequence of whole-Earth decompression dynamics:\nnamely, a specific, dominant, non-erosion, underlying initiation-mechanism\nprecursor for submarine canyons that follows as a direct consequence of Earth's\nearly origin as a Jupiter-like gas-giant.\n", " An interpretation of non-relativistic quantum mechanics is presented in the\nspirit of Erwin Madelung's hydrodynamic formulation of QM and Louis de\nBroglie's and David Bohm's pilot wave models. The aims of the approach are as\nfollows: 1) to have a clear ontology for QM, 2) to describe QM in a causal way,\n3) to get rid of the wave-particle dualism in pilot wave theories, 4) to\nprovide a theoretical framework for describing creation and annihilation of\nparticles, and 5) to provide a possible connection between particle QM and\nvirtual particles in QFT. These goals are achieved, if the wave function is\nreplaced by a fluid of so called virtual particles. It is also assumed that in\nthis fluid of virtual particles exist a few real particles and that only these\nreal particles can be directly observed. This has relevance for the measurement\nproblem in QM and it is found that quantum probabilities arise in a very\nnatural way from the structure of the theory. The model presented here is very\nsimilar to a recent computational model of quantum physics and recent Bohmian\nmodels of QFT.\n", " We present a full theoretical and experimental study of the dynamics and\nenergy distribution of non-equilibrium quasiparticles in superconducting tunnel\njunctions (STJs). STJs are often used for single-photon spectrometers, where\nthe numbers of quasiparticles excited by a photon provide a measure of the\nphoton energy. The magnitude and fluctuations of the signal current in STJ\ndetectors are in large part determined by the quasiparticle dynamics and energy\ndistribution during the detection process. We use this as motivation to study\nthe transport and energy distribution of non-equilibrium quasiparticles excited\nby x-ray photons in a lateral, imaging junction configuration. We present a\nfull numerical model for the tunneling current of the major physical processes\nwhich determine the signal. We find that a diffusion framework models the\nquasiparticle dynamics well and that excited quasiparticles do not equilibrate\nto the lattice temperature during the timescales for tunneling. We extract\nphysical timescales from the measured data, make comparisons with existing\ntheories, and comment on implications for superconducting mesoscopic systems\nand single-photon detectors.\n", " We present a general and modular algorithmic framework for path planning of\nrobots. Our framework combines geometric methods for exact and complete\nanalysis of low-dimensional configuration spaces, together with practical,\nconsiderably simpler sampling-based approaches that are appropriate for higher\ndimensions. In order to facilitate the transfer of advanced geometric\nalgorithms into practical use, we suggest taking samples that are entire\nlow-dimensional manifolds of the configuration space that capture the\nconnectivity of the configuration space much better than isolated point\nsamples. Geometric algorithms for analysis of low-dimensional manifolds then\nprovide powerful primitive operations. The modular design of the framework\nenables independent optimization of each modular component. Indeed, we have\ndeveloped, implemented and optimized a primitive operation for complete and\nexact combinatorial analysis of a certain set of manifolds, using arrangements\nof curves of rational functions and concepts of generic programming. This in\nturn enabled us to implement our framework for the concrete case of a polygonal\nrobot translating and rotating amidst polygonal obstacles. We demonstrate that\nthe integration of several carefully engineered components leads to significant\nspeedup over the popular PRM sampling-based algorithm, which represents the\nmore simplistic approach that is prevalent in practice. We foresee possible\nextensions of our framework to solving high-dimensional problems beyond motion\nplanning.\n", " Any organism is embedded in an environment that changes over time. The\ntimescale for and statistics of environmental change, the precision with which\nthe organism can detect its environment, and the costs and benefits of\nparticular protein expression levels all will affect the suitability of\ndifferent strategies-such as constitutive expression or graded response-for\nregulating protein levels in response to environmental inputs. We propose a\ngeneral framework-here specifically applied to the enzymatic regulation of\nmetabolism in response to changing concentrations of a basic nutrient-to\npredict the optimal regulatory strategy given the statistics of fluctuations in\nthe environment and measurement apparatus, respectively, and the costs\nassociated with enzyme production. We use this framework to address three\nfundamental questions: (i) when a cell should prefer thresholding to a graded\nresponse; (ii) when there is a fitness advantage to implementing a Bayesian\ndecision rule; and (iii) when retaining memory of the past provides a selective\nadvantage. We specifically find that: (i) relative convexity of enzyme\nexpression cost and benefit influences the fitness of thresholding or graded\nresponses; (ii) intermediate levels of measurement uncertainty call for a\nsophisticated Bayesian decision rule; and (iii) in dynamic contexts,\nintermediate levels of uncertainty call for retaining memory of the past.\nStatistical properties of the environment, such as variability and correlation\ntimes, set optimal biochemical parameters, such as thresholds and decay rates\nin signaling pathways. Our framework provides a theoretical basis for\ninterpreting molecular signal processing algorithms and a classification scheme\nthat organizes known regulatory strategies and may help conceptualize\nheretofore unknown ones.\n", " Variational inference is a popular technique to approximate a possibly\nintractable Bayesian posterior with a more tractable one. Recently, boosting\nvariational inference has been proposed as a new paradigm to approximate the\nposterior by a mixture of densities by greedily adding components to the\nmixture. However, as is the case with many other variational inference\nalgorithms, its theoretical properties have not been studied. In the present\nwork, we study the convergence properties of this approach from a modern\noptimization viewpoint by establishing connections to the classic Frank-Wolfe\nalgorithm. Our analyses yields novel theoretical insights regarding the\nsufficient conditions for convergence, explicit rates, and algorithmic\nsimplifications. Since a lot of focus in previous works for variational\ninference has been on tractability, our work is especially important as a much\nneeded attempt to bridge the gap between probabilistic models and their\ncorresponding theoretical properties.\n", " We develop a theory of discrete directional Gabor frames for functions\ndefined on the $d$-dimensional Euclidean space. Our construction incorporates\nthe concept of ridge functions into the theory of isotropic Gabor systems, in\norder to develop an anisotropic Gabor system with strong directional\nsensitivity. We present sufficient conditions on a window function $g$ and a\nsampling set $\\Lambda$ for the corresponding directional Gabor system\n$\\{g_{m,t,u}\\}_{(m,t,u)\\in\\Lambda}$ to form a discrete frame. Explicit\nestimates on the frame bounds are developed. A numerical implementation of our\nscheme is also presented, and is shown to perform competitively in compression\nand denoising against state-of-the-art anisotropic methods.\n", " Bernstein polynomials and B\\'ezier curves play an important role in\ncomputer-aided geometric design and numerical analysis, and their study relates\nto mathematical fields such as abstract algebra, algebraic geometry and\nprobability theory. We describe a theoretical framework that incorporates the\ndifferent aspects of the Bernstein-B\\'ezier theory, based on concepts from\ntheoretical physics. We relate B\\'ezier curves to the theory of angular\nmomentum in both classical and quantum mechanics, and describe physical\nanalogues of various properties of B\\'ezier curves -- such as their connection\nwith polar forms -- in the context of quantum spin systems. This previously\nunexplored relationship between geometric design and theoretical physics is\nestablished using the mathematical theory of Hamiltonian mechanics and\ngeometric quantization. An alternative description of spin systems in terms of\nharmonic oscillators serves as a physical analogue of P\\'olya's urn models for\nB\\'ezier curves. We relate harmonic oscillators to Poisson curves and the\nanalytical blossom as well. We present an overview of the relevant mathematical\nand physical concepts, and discuss opportunities for further research.\n", " Persistence diagrams are one of the main tools in the field of Topological\nData Analysis (TDA). They contain fruitful information about the shape of data.\nThe use of machine learning algorithms on the space of persistence diagrams\nproves to be challenging as the space lacks an inner product. For that reason,\ntransforming these diagrams in a way that is compatible with machine learning\nis an important topic currently researched in TDA. In this paper, our main\ncontribution consists of three components. First, we develop a general and\nunifying framework of vectorizing diagrams that we call the \\textit{Persistence\nCurves} (PCs), and show that several well-known summaries, such as Persistence\nLandscapes, fall under the PC framework. Second, we propose several new\nsummaries based on PC framework and provide a theoretical foundation for their\nstability analysis. Finally, we apply proposed PCs to two\napplications---texture classification and determining the parameters of a\ndiscrete dynamical system; their performances are competitive with other TDA\nmethods.\n", " Minkowski space serves as a framework for the theoretical constructions that\ndeal with manifestations of relativistic effects in physical phenomena. But\nneither Minkowski himself nor the subsequent developers of the relativity\ntheory have provided a reasonable rationale for this mathematical construct. In\nphysics, such a rationale should show lower-level statements that determine\nwhere the proposed mathematical structure is applicable and yield formal\npremises for proving its existence.\n The above failure has apparently been due to the features of the adopted\nformalism based on the unjustifiably exclusive use of coordinates in the\ntheoretical analysis of physical phenomena, which ignores the necessity of\nhaving physical grounds for mathematical concepts. In particular, the use of a\ncoordinate transformation between two inertial reference frames makes the\nconsideration so cumbersome that it appears useless for solving the fundamental\nproblems of physical theory, including the question of whether Minkowski space\nexists.\n In contrast, a straightforward calculation proves that the transformation of\nthe time and the position vector of a physical event between two physical\nspaces establishes an equivalence relation between pairs made of these\nvariables. This means the existence of Minkowski space and shows that the\npremises for its proof are the same as for the coordinate-free derivation of\nbasic effects of the special relativity theory: the use of Einsteinian time\nvariable and motions of particles able to interact with each other and\nelectromagnetic field over a short spatial range only.\n The high degeneracy of free motions of point particles, together with the\nintricacy of the above mentioned calculation, suggests that a further\ngeneralization of Minkowski space is beyond belief, so that the modification or\neven the abandonment of the concept of spacetime seems quite natural.\n", " Gravitational-wave experiments with interferometers and with resonant masses\ncan search for stochastic backgrounds of gravitational waves of cosmological\norigin. We review both experimental and theoretical aspects of the search for\nthese backgrounds. We give a pedagogical derivation of the various relations\nthat characterize the response of a detector to a stochastic background. We\ndiscuss the sensitivities of the large interferometers under constructions\n(LIGO, VIRGO, GEO600, TAMA300, AIGO) or planned (Avdanced LIGO, LISA) and of\nthe presently operating resonant bars, and we give the sensitivities for\nvarious two-detectors correlations. We examine the existing limits on the\nenergy density in gravitational waves from nucleosynthesis, COBE and pulsars,\nand their effects on theoretical predictions. We discuss general theoretical\nprinciples for order-of-magnitude estimates of cosmological production\nmechanisms, and then we turn to specific theoretical predictions from\ninflation, string cosmology, phase transitions, cosmic strings and other\nmechanisms. We finally compare with the stochastic backgrounds of astrophysical\norigin.\n", " Many biological processes discriminate between correct and incorrect\nsubstrates through the kinetic proofreading mechanism which enables lower error\nat the cost of higher energy dissipation. Elucidating physicochemical\nconstraints for global minimization of dissipation and error is important for\nunderstanding enzyme evolution. Here, we identify theoretically a fundamental\nerror-cost bound which tightly constrains the performance of proofreading\nnetworks under any parameter variations preserving the rate discrimination\nbetween substrates. The bound is kinetically controlled, i.e. completely\ndetermined by the difference between the transition state energies on the\nunderlying free energy landscape. The importance of the bound is analyzed for\nthree biological processes. DNA replication by T7 DNA polymerase is shown to be\nnearly optimized, i.e. its kinetic parameters place it in the immediate\nproximity of the error-cost bound. The isoleucyl-tRNA synthetase (IleRS) of E.\ncoli also operates close to the bound, but further optimization is prevented by\nthe need for reaction speed. In contrast, E. coli ribosome operates in a\nhigh-dissipation regime, potentially in order to speed up protein production.\nTogether, these findings establish a fundamental error-dissipation relation in\nbiological proofreading networks and provide a theoretical framework for\nstudying error-dissipation trade-off in other systems with biological\ndiscrimination.\n", " In this paper, we present a framework for getting a series of exact vacuum\nsolutions to the Einstein equation. This procedure of resolution is based on a\ncanonical form of the metric. According to this procedure, the Einstein\nequation can be reduced to some 2-dimensional Laplace-like equations or\nrotation and divergence equations, which are much convenient for the\nresolution.\n", " We give explicit theoretical and heuristical bounds for how big does a data\nset sampled from a reach-1 submanifold M of euclidian space need to be, to be\nable to estimate the dimension of M with 90% confidence.\n", " Nonuniformities in the imaging characteristics of modern image sensors are a\nprimary factor in the push to develop a pixel-level generalization of the\nphoton transfer characterization method. In this paper, we seek to develop a\nbody of theoretical results leading toward a comprehensive approach for\ntackling the biggest obstacle in the way of this goal: a means of pixel-level\nconversion gain estimation. This is accomplished by developing an estimator for\nthe reciprocal-difference of normal variances and then using this to construct\na novel estimator of the conversion gain. The first two moments of this\nestimator are derived and used to construct exact and approximate confidence\nintervals for its absolute relative bias and absolute coefficient of variation,\nrespectively. A means of approximating and computing optimal sample sizes are\nalso discussed and used to demonstrate the process of pixel-level conversion\ngain estimation for a real image sensor.\n", " Automated tutoring systems offer the flexibility and scalability necessary to\nfacilitate the provision of high quality and universally accessible programming\neducation. In order to realise the full potential of these systems, recent work\nhas proposed a diverse range of techniques for automatically generating hints\nto assist students with programming exercises. This paper integrates these\napparently disparate approaches into a coherent whole. Specifically, it\nemphasises that all hint techniques can be understood as a series of simpler\ncomponents with similar properties. Using this insight, it presents a simple\nframework for describing such techniques, the Hint Iteration by Narrow-down and\nTransformation Steps (HINTS) framework, and it surveys recent work in the\ncontext of this framework. It discusses important implications of the survey\nand framework, including the need to further develop evaluation methods and the\nimportance of considering hint technique components when designing,\ncommunicating and evaluating hint systems. Ultimately, this paper is designed\nto facilitate future opportunities for the development, extension and\ncomparison of automated programming hint techniques in order to maximise their\neducational potential.\n", " Conjecturing and theorem proving are activities at the center of mathematical\npractice and are difficult to separate. In this paper, we propose a framework\nfor completing incomplete conjectures and incomplete proofs. The framework can\nturn a conjecture with missing assumptions and with an under-specified goal\ninto a proper theorem. Also, the proposed framework can help in completing a\nproof sketch into a human-readable and machine-checkable proof. Our approach is\nfocused on synthetic geometry, and uses coherent logic and constraint solving.\nThe proposed approach is uniform for all three kinds of tasks, flexible and, to\nour knowledge, unique such approach.\n", " We describe a general probabilistic framework to address a variety of\nFrechet-distance optimization problems. Specifically, we are interested in\nfinding minimal bottleneck-paths in $d$-dimensional Euclidean space between\ngiven start and goal points, namely paths that minimize the maximal value over\na continuous cost map. We present an efficient and simple sampling-based\nframework for this problem, which is inspired by, and draws ideas from,\ntechniques for robot motion planning. We extend the framework to handle not\nonly standard bottleneck pathfinding, but also the more demanding case, where\nthe path needs to be monotone in all dimensions. Finally, we provide\nexperimental results of the framework on several types of problems.\n", " Real-valued functions on geometric data -- such as node attributes on a graph\n-- can be optimized using descriptors from persistent homology, allowing the\nuser to incorporate topological terms in the loss function. When optimizing a\nsingle real-valued function (the one-parameter setting), there is a canonical\nchoice of descriptor for persistent homology: the barcode. The operation\nmapping a real-valued function to its barcode is differentiable almost\neverywhere, and the convergence of gradient descent for losses using barcodes\nis relatively well understood. When optimizing a vector-valued function (the\nmultiparameter setting), there is no unique choice of descriptor for\nmultiparameter persistent homology, and many distinct descriptors have been\nproposed. This calls for the development of a general framework for\ndifferentiability and optimization that applies to a wide range of\nmultiparameter homological descriptors. In this article, we develop such a\nframework and show that it encompasses well-known descriptors of different\nflavors, such as signed barcodes and the multiparameter persistence landscape.\nWe complement the theory with numerical experiments supporting the idea that\noptimizing multiparameter homological descriptors can lead to improved\nperformances compared to optimizing one-parameter descriptors, even when using\nthe simplest and most efficiently computable multiparameter descriptors.\n", " Convolutional Neural Networks (CNN) are the state-of-the-art in the field of\nvisual computing. However, a major problem with CNNs is the large number of\nfloating point operations (FLOPs) required to perform convolutions for large\ninputs. When considering the application of CNNs to video data, convolutional\nfilters become even more complex due to the extra temporal dimension. This\nleads to problems when respective applications are to be deployed on mobile\ndevices, such as smart phones, tablets, micro-controllers or similar,\nindicating less computational power.\n Kim et al. (2016) proposed using a Tucker-decomposition to compress the\nconvolutional kernel of a pre-trained network for images in order to reduce the\ncomplexity of the network, i.e. the number of FLOPs. In this paper, we\ngeneralize the aforementioned method for application to videos (and other 3D\nsignals) and evaluate the proposed method on a modified version of the THETIS\ndata set, which contains videos of individuals performing tennis shots. We show\nthat the compressed network reaches comparable accuracy, while indicating a\nmemory compression by a factor of 51. However, the actual computational\nspeed-up (factor 1.4) does not meet our theoretically derived expectation\n(factor 6).\n", " One noticed a numerical coincidence between the Pioneer spacecrafts\ndeceleration anomaly and $(\\gamma-1)$, with $\\gamma$ the Lorentz factor. The\nmatch is not only for distances larger than 20 AU, but even for the observed\nslop between 10 and 20 AU. Such numerical link may eventually lead to a\nscientific hypothesis for future theoretical investigations.\n", " In this paper, we introduce the notion of DTM-signature, a measure on R +\nthat can be associated to any metric-measure space. This signature is based on\nthe distance to a measure (DTM) introduced by Chazal, Cohen-Steiner and\nM\\'erigot. It leads to a pseudo-metric between metric-measure spaces,\nupper-bounded by the Gromov-Wasserstein distance. Under some geometric\nassumptions, we derive lower bounds for this pseudo-metric. Given two\nN-samples, we also build an asymptotic statistical test based on the\nDTM-signature, to reject the hypothesis of equality of the two underlying\nmetric-measure spaces, up to a measure-preserving isometry. We give strong\ntheoretical justifications for this test and propose an algorithm for its\nimplementation.\n", " We give a general approach to infinite dimensional non-Gaussian Analysis for\nmeasures which need not have a logarithmic derivative. This framework also\nincludes the possibility to handle measures of Poisson type.\n", " We study the problem of discrete geometric packing. Here, given weighted\nregions (say in the plane) and points (with capacities), one has to pick a\nmaximum weight subset of the regions such that no point is covered more than\nits capacity. We provide a general framework and an algorithm for approximating\nthe optimal solution for packing in hypergraphs arising out of such geometric\nsettings. Using this framework we get a flotilla of results on this problem\n(and also on its dual, where one wants to pick a maximum weight subset of the\npoints when the regions have capacities). For example, for the case of fat\ntriangles of similar size, we show an O(1)-approximation and prove that no\n\\PTAS is possible.\n", " We proposed a new criterion \\textit{noise-stability}, which revised the\nclassical rigidity theory, for evaluation of MDS algorithms which can\ntruthfully represent the fidelity of global structure reconstruction; then we\nproved the noise-stability of the cMDS algorithm in generic conditions, which\nprovides a rigorous theoretical guarantee for the precision and theoretical\nbounds for Euclidean embedding and its application in fields including wireless\nsensor network localization and satellite positioning.\n Furthermore, we looked into previous work about minimum-cost globally rigid\nspanning subgraph, and proposed an algorithm to construct a minimum-cost\nnoise-stable spanning graph in the Euclidean space, which enabled reliable\nlocalization on sparse graphs of noisy distance constraints with linear numbers\nof edges and sublinear costs in total edge lengths. Additionally, this\nalgorithm also suggests a scheme to reconstruct point clouds from pairwise\ndistances at a minimum of $O(n)$ time complexity, down from $O(n^3)$ for cMDS.\n", " This is a preliminary theoretical discussion on the computational\nrequirements of the state of the art smoothed particle hydrodynamics (SPH) from\nthe optics of pattern recognition and artificial intelligence. It is pointed\nout in the present paper that, when including anisotropy detection to improve\nresolution on shock layer, SPH is a very peculiar case of unsupervised machine\nlearning. On the other hand, the free particle nature of SPH opens an\nopportunity for artificial intelligence to study particles as agents acting in\na collaborative framework in which the timed outcomes of a fluid simulation\nforms a large knowledge base, which might be very attractive in computational\nastrophysics phenomenological problems like self-propagating star formation.\n", " Prediction+optimization is a common real-world paradigm where we have to\npredict problem parameters before solving the optimization problem. However,\nthe criteria by which the prediction model is trained are often inconsistent\nwith the goal of the downstream optimization problem. Recently,\ndecision-focused prediction approaches, such as SPO+ and direct optimization,\nhave been proposed to fill this gap. However, they cannot directly handle the\nsoft constraints with the $max$ operator required in many real-world\nobjectives. This paper proposes a novel analytically differentiable surrogate\nobjective framework for real-world linear and semi-definite negative quadratic\nprogramming problems with soft linear and non-negative hard constraints. This\nframework gives the theoretical bounds on constraints' multipliers, and derives\nthe closed-form solution with respect to predictive parameters and thus\ngradients for any variable in the problem. We evaluate our method in three\napplications extended with soft constraints: synthetic linear programming,\nportfolio optimization, and resource provisioning, demonstrating that our\nmethod outperforms traditional two-staged methods and other decision-focused\napproaches.\n", " We provide a rigorous theoretical basis for the artificial magnetic activity\nof meta-materials near resonances. Our approach is a renormalization-based\nscheme that authorises a completely general theory. The major result is an\nexplicit expression of the effective permeability, in terms of resonant\nfrequencies. The theoretical results are checked numerically and we give\napplications of our theory to Left-Handed Media and to the solution of the\nPokrovski-Efros paradox.\n", " Atomistic simulations provide a meaningful way to determine the\nphysico-chemical properties of liquids in a consistent theoretical framework.\nThis approach takes on particular usefulness for the study of molten\ncarbonates, in a context where thermodynamic and transport data are crucially\nneeded over a large domain of temperatures and pressures (to ascertain the role\nof these melts in geochemical processes) but are very scarce in the literature,\nespecially for the calco-magnesian compositions prevailing in the Earth's\nmantle. Following our work on Li2CO3-Na2CO3-K2CO3 melts, we extend our force\nfield to incorporate Ca and Mg components. The empirical interaction potentials\nare benchmarked on the density data available in the experimental literature\n(for the crystals and the K2Ca(CO3)2 melt) and on the liquid structure issued\nfrom ab initio molecular dynamics simulations. Molecular dynamics simulations\nare then performed to study the thermodynamics, the microscopic structure, the\ndiffusion coefficients, the electrical conductivity and the viscosity of molten\nCa, Mg-bearing carbonates up to 2073 K and 15 GPa. Additionally, the equation\nof state of a Na-Ca-K mixture representative of the lavas emitted at Ol Doinyo\nLengai (Tanzania) is evaluated. The overall agreement between the MD results\nand the existing experimental data is very satisfying and provides evidence for\nthe ability of the force field to accurately model any\nMgCO3-CaCO3-Li2CO3-Na2CO3-K2CO3 melt over a large T - P range. Moreover it is\nthe first report of a force field allowing to study the transport properties of\nmolten magnesite (MgCO3) and molten dolomite (CaMg(CO3)2).\n", " We study a noncommutative theory of gravity in the framework of torsional\nspacetime. This theory is based on a Lagrangian obtained by applying the\ntechnique of dimensional reduction of noncommutative gauge theory and that the\nyielded diffeomorphism invariant field theory can be made equivalent to a\nteleparallel formulation of gravity. Field equations are derived in the\nframework of teleparallel gravity through Weitzenbock geometry. We solve these\nfield equations by considering a mass that is distributed spherically\nsymmetrically in a stationary static spacetime in order to obtain a\nnoncommutative line element.This new line element interestingly reaffirms the\ncoherent state theory for a noncommutative Schwarzschild black hole. For the\nfirst time, we derive the Newtonian gravitational force equation in the\ncommutative relativity framework, and this result could provide the possibility\nto investigate examples in various topics in quantum and ordinary theories of\ngravity.\n", " Stochastic Boolean networks, or more generally, stochastic discrete networks,\nare an important class of computational models for molecular interaction\nnetworks. The stochasticity stems from the updating schedule. Standard updating\nschedules include the synchronous update, where all the nodes are updated at\nthe same time, and the asynchronous update where a random node is updated at\neach time step. The former produces a deterministic dynamics while the latter a\nstochastic dynamics. A more general stochastic setting considers propensity\nparameters for updating each node. Stochastic Discrete Dynamical Systems (SDDS)\nis a modeling framework that considers two propensity parameters for updating\neach node and uses one when the update has a positive impact on the variable,\nthat is, when the update causes the variable to increase its value, and uses\nthe other when the update has a negative impact, that is, when the update\ncauses it to decrease its value. This framework offers additional features for\nsimulations but also adds a complexity in parameter estimation of the\npropensities. This paper presents a method for estimating the propensity\nparameters for SDDS. The method is based on adding noise to the system using\nthe Google PageRank approach to make the system ergodic and thus guaranteeing\nthe existence of a stationary distribution. Then with the use of a genetic\nalgorithm, the propensity parameters are estimated. Approximation techniques\nthat make the search algorithms efficient are also presented and Matlab/Octave\ncode to test the algorithms are available at\nhttp://www.ms.uky.edu/~dmu228/GeneticAlg/Code.html.\n" ]

[ " Mobile Edge Computing (MEC) broadens the scope of computation and storage\nbeyond the central network, incorporating edge nodes close to end devices. This\nexpansion facilitates the implementation of large-scale \"connected things\"\nwithin edge networks. The advent of applications necessitating real-time,\nhigh-quality service presents several challenges, such as low latency, high\ndata rate, reliability, efficiency, and security, all of which demand\nresolution. The incorporation of reinforcement learning (RL) methodologies\nwithin MEC networks promotes a deeper understanding of mobile user behaviors\nand network dynamics, thereby optimizing resource use in computing and\ncommunication processes. This paper offers an exhaustive survey of RL\napplications in MEC networks, initially presenting an overview of RL from its\nfundamental principles to the latest advanced frameworks. Furthermore, it\noutlines various RL strategies employed in offloading, caching, and\ncommunication within MEC networks. Finally, it explores open issues linked with\nsoftware and hardware platforms, representation, RL robustness, safe RL,\nlarge-scale scheduling, generalization, security, and privacy. The paper\nproposes specific RL techniques to mitigate these issues and provides insights\ninto their practical applications.\n", " In the case of the need of extraordinary security,\nKirchhoff-loop-Johnson-(like)-noise ciphers can easily be integrated on\nexisting types of digital chips in order to provide secure data communication\nbetween hardware processors, memory chips, hard disks and other units within a\ncomputer or other data processor system. The secure key exchange can take place\nat the very first run and the system can renew the key later at random times\nwith an authenticated fashion to prohibit man-in-the-middle attack. The key can\nbe stored in flash memories within the communicating chip units at hidden\nrandom addresses among other random bits that are continuously generated by the\nsecure line but are never actually used. Thus, even if the system is\ndisassembled, and the eavesdropper can have direct access to the communication\nlines between the units, or even if she is trying to use a man-in-the-middle\nattack, no information can be extracted. The only way to break the code is to\nlearn the chip structure, to understand the machine code program and to read\nout the information during running by accessing the proper internal ports of\nthe working chips. However such an attack needs extraordinary resources and\neven that can be prohibited by a password lockout. The unconditional security\nof commercial algorithms against piracy can be provided in a similar way.\n", " Due to the rapid spread of rumors on social media, rumor detection has become\nan extremely important challenge. Recently, numerous rumor detection models\nwhich utilize textual information and the propagation structure of events have\nbeen proposed. However, these methods overlook the importance of semantic\nevolvement information of event in propagation process, which is often\nchallenging to be truly learned in supervised training paradigms and\ntraditional rumor detection methods. To address this issue, we propose a novel\nsemantic evolvement enhanced Graph Autoencoder for Rumor Detection (GARD) model\nin this paper. The model learns semantic evolvement information of events by\ncapturing local semantic changes and global semantic evolvement information\nthrough specific graph autoencoder and reconstruction strategies. By combining\nsemantic evolvement information and propagation structure information, the\nmodel achieves a comprehensive understanding of event propagation and perform\naccurate and robust detection, while also detecting rumors earlier by capturing\nsemantic evolvement information in the early stages. Moreover, in order to\nenhance the model's ability to learn the distinct patterns of rumors and\nnon-rumors, we introduce a uniformity regularizer to further improve the\nmodel's performance. Experimental results on three public benchmark datasets\nconfirm the superiority of our GARD method over the state-of-the-art approaches\nin both overall performance and early rumor detection.\n", " Deep neural networks are known to have security issues. One particular threat\nis the Trojan attack. It occurs when the attackers stealthily manipulate the\nmodel's behavior through Trojaned training samples, which can later be\nexploited.\n Guided by basic neuroscientific principles we discover subtle -- yet critical\n-- structural deviation characterizing Trojaned models. In our analysis we use\ntopological tools. They allow us to model high-order dependencies in the\nnetworks, robustly compare different networks, and localize structural\nabnormalities. One interesting observation is that Trojaned models develop\nshort-cuts from input to output layers.\n Inspired by these observations, we devise a strategy for robust detection of\nTrojaned models. Compared to standard baselines it displays better performance\non multiple benchmarks.\n", " The next leap on the internet has already started as Semantic Web. At its\ncore, Semantic Web transforms the document oriented web to a data oriented web\nenriched with semantics embedded as metadata. This change in perspective\ntowards the web offers numerous benefits for vast amount of data intensive\nindustries that are bound to the web and its related applications. The\nindustries are diverse as they range from Oil & Gas exploration to the\ninvestigative journalism, and everything in between. This paper discusses eight\ndifferent industries which currently reap the benefits of Semantic Web. The\npaper also offers a future outlook into Semantic Web applications and discusses\nthe areas in which Semantic Web would play a key role in the future.\n", " In recent years, JavaScript has become the most widely used programming\nlanguage, especially in web development. However, writing secure JavaScript\ncode is not trivial, and programmers often make mistakes that lead to security\nvulnerabilities in web applications. Large Language Models (LLMs) have\ndemonstrated substantial advancements across multiple domains, and their\nevolving capabilities indicate their potential for automatic code generation\nbased on a required specification, including automatic bug fixing. In this\nstudy, we explore the accuracy of LLMs, namely ChatGPT and Bard, in finding and\nfixing security vulnerabilities in JavaScript programs. We also investigate the\nimpact of context in a prompt on directing LLMs to produce a correct patch of\nvulnerable JavaScript code. Our experiments on real-world software\nvulnerabilities show that while LLMs are promising in automatic program repair\nof JavaScript code, achieving a correct bug fix often requires an appropriate\namount of context in the prompt.\n", " In this paper, we consider the statistical analysis of a protein interaction\nnetwork. We propose a Bayesian model that uses a hierarchy of probabilistic\nassumptions about the way proteins interact with one another in order to: (i)\nidentify the number of non-observable functional modules; (ii) estimate the\ndegree of membership of proteins to modules; and (iii) estimate typical\ninteraction patterns among the functional modules themselves. Our model\ndescribes large amount of (relational) data using a relatively small set of\nparameters that we can reliably estimate with an efficient inference algorithm.\nWe apply our methodology to data on protein-to-protein interactions in\nsaccharomyces cerevisiae to reveal proteins' diverse functional roles. The case\nstudy provides the basis for an overview of which scientific questions can be\naddressed using our methods, and for a discussion of technical issues.\n", " We address, through the automated reasoning tools in GeoGebra Discovery, a\nproblem from a regional phase of the Austrian Mathematics Olympiad 2023. Trying\nto solve this problem gives rise to four different kind of feedback: the almost\ninstantaneous, automated solution of the proposed problem; the measure of its\ncomplexity, according to some recent proposals; the automated discovery of a\ngeneralization of the given assertion, showing that the same statement is true\nover more general polygons than those mentioned in the problem; and the\ndifficulties associated to the analysis of the surprising and involved high\nnumber of degenerate cases that appear when using the LocusEquation command in\nthis problem. In our communication we will describe and reflect on these\ndiverse issues, enhancing its exemplar role for showing some of the advantages,\nproblems, and current fields of development of GeoGebra Discovery.\n", " Traffic signal control (TSC) is a challenging problem within intelligent\ntransportation systems and has been tackled using multi-agent reinforcement\nlearning (MARL). While centralized approaches are often infeasible for\nlarge-scale TSC problems, decentralized approaches provide scalability but\nintroduce new challenges, such as partial observability. Communication plays a\ncritical role in decentralized MARL, as agents must learn to exchange\ninformation using messages to better understand the system and achieve\neffective coordination. Deep MARL has been used to enable inter-agent\ncommunication by learning communication protocols in a differentiable manner.\nHowever, many deep MARL communication frameworks proposed for TSC allow agents\nto communicate with all other agents at all times, which can add to the\nexisting noise in the system and degrade overall performance. In this study, we\npropose a communication-based MARL framework for large-scale TSC. Our framework\nallows each agent to learn a communication policy that dictates \"which\" part of\nthe message is sent \"to whom\". In essence, our framework enables agents to\nselectively choose the recipients of their messages and exchange variable\nlength messages with them. This results in a decentralized and flexible\ncommunication mechanism in which agents can effectively use the communication\nchannel only when necessary. We designed two networks, a synthetic $4 \\times 4$\ngrid network and a real-world network based on the Pasubio neighborhood in\nBologna. Our framework achieved the lowest network congestion compared to\nrelated methods, with agents utilizing $\\sim 47-65 \\%$ of the communication\nchannel. Ablation studies further demonstrated the effectiveness of the\ncommunication policies learned within our framework.\n", " We propose a targeted communication architecture for multi-agent\nreinforcement learning, where agents learn both what messages to send and whom\nto address them to while performing cooperative tasks in partially-observable\nenvironments. This targeting behavior is learnt solely from downstream\ntask-specific reward without any communication supervision. We additionally\naugment this with a multi-round communication approach where agents coordinate\nvia multiple rounds of communication before taking actions in the environment.\nWe evaluate our approach on a diverse set of cooperative multi-agent tasks, of\nvarying difficulties, with varying number of agents, in a variety of\nenvironments ranging from 2D grid layouts of shapes and simulated traffic\njunctions to 3D indoor environments, and demonstrate the benefits of targeted\nand multi-round communication. Moreover, we show that the targeted\ncommunication strategies learned by agents are interpretable and intuitive.\nFinally, we show that our architecture can be easily extended to mixed and\ncompetitive environments, leading to improved performance and sample complexity\nover recent state-of-the-art approaches.\n", " We present a method that detects boundaries of parts in 3D shapes represented\nas point clouds. Our method is based on a graph convolutional network\narchitecture that outputs a probability for a point to lie in an area that\nseparates two or more parts in a 3D shape. Our boundary detector is quite\ngeneric: it can be trained to localize boundaries of semantic parts or\ngeometric primitives commonly used in 3D modeling. Our experiments demonstrate\nthat our method can extract more accurate boundaries that are closer to\nground-truth ones compared to alternatives. We also demonstrate an application\nof our network to fine-grained semantic shape segmentation, where we also show\nimprovements in terms of part labeling performance.\n", " The rapid advancements in large language models (LLMs) have opened new\navenues across various fields, including cybersecurity, which faces an\never-evolving threat landscape and need for innovative technologies. Despite\ninitial explorations into the application of LLMs in cybersecurity, there is a\nlack of a comprehensive overview of this research area. This paper bridge this\ngap by providing a systematic literature review, encompassing an analysis of\nover 180 works, spanning across 25 LLMs and more than 10 downstream scenarios.\nOur comprehensive overview addresses three critical research questions: the\nconstruction of cybersecurity-oriented LLMs, LLMs' applications in various\ncybersecurity tasks, and the existing challenges and further research in this\narea. This study aims to shed light on the extensive potential of LLMs in\nenhancing cybersecurity practices, and serve as a valuable resource for\napplying LLMs in this doamin. We also maintain and regularly updated list of\npractical guides on LLMs for cybersecurity at\nhttps://github.com/tmylla/Awesome-LLM4Cybersecurity.\n", " We give a brief outline of biography, review 3 of his works on the use of\nconfluent hypergeometric function for description asymmetric relaxation\nspectrum and provide an overview of the Soviet works of the past (20th) that\nare in tune with this topic.\n", " A parallel algorithm has perfect strong scaling if its running time on P\nprocessors is linear in 1/P, including all communication costs.\nDistributed-memory parallel algorithms for matrix multiplication with perfect\nstrong scaling have only recently been found. One is based on classical matrix\nmultiplication (Solomonik and Demmel, 2011), and one is based on Strassen's\nfast matrix multiplication (Ballard, Demmel, Holtz, Lipshitz, and Schwartz,\n2012). Both algorithms scale perfectly, but only up to some number of\nprocessors where the inter-processor communication no longer scales.\n We obtain a memory-independent communication cost lower bound on classical\nand Strassen-based distributed-memory matrix multiplication algorithms. These\nbounds imply that no classical or Strassen-based parallel matrix multiplication\nalgorithm can strongly scale perfectly beyond the ranges already attained by\nthe two parallel algorithms mentioned above. The memory-independent bounds and\nthe strong scaling bounds generalize to other algorithms.\n", " Anomaly detection is a critical challenge across various research domains,\naiming to identify instances that deviate from normal data distributions. This\npaper explores the application of Generative Adversarial Networks (GANs) in\nfraud detection, comparing their advantages with traditional methods. GANs, a\ntype of Artificial Neural Network (ANN), have shown promise in modeling complex\ndata distributions, making them effective tools for anomaly detection. The\npaper systematically describes the principles of GANs and their derivative\nmodels, emphasizing their application in fraud detection across different\ndatasets. And by building a collection of adversarial verification graphs, we\nwill effectively prevent fraud caused by bots or automated systems and ensure\nthat the users in the transaction are real. The objective of the experiment is\nto design and implement a fake face verification code and fraud detection\nsystem based on Generative Adversarial network (GANs) algorithm to enhance the\nsecurity of the transaction process.The study demonstrates the potential of\nGANs in enhancing transaction security through deep learning techniques.\n", " Current research on Internet of Things (IoT) mainly focuses on how to enable\ngeneral objects to see, hear, and smell the physical world for themselves, and\nmake them connected to share the observations. In this paper, we argue that\nonly connected is not enough, beyond that, general objects should have the\ncapability to learn, think, and understand both physical and social worlds by\nthemselves. This practical need impels us to develop a new paradigm, named\nCognitive Internet of Things (CIoT), to empower the current IoT with a `brain'\nfor high-level intelligence. Specifically, we first present a comprehensive\ndefinition for CIoT, primarily inspired by the effectiveness of human\ncognition. Then, we propose an operational framework of CIoT, which mainly\ncharacterizes the interactions among five fundamental cognitive tasks:\nperception-action cycle, massive data analytics, semantic derivation and\nknowledge discovery, intelligent decision-making, and on-demand service\nprovisioning. Furthermore, we provide a systematic tutorial on key enabling\ntechniques involved in the cognitive tasks. In addition, we also discuss the\ndesign of proper performance metrics on evaluating the enabling techniques.\nLast but not least, we present the research challenges and open issues ahead.\nBuilding on the present work and potentially fruitful future studies, CIoT has\nthe capability to bridge the physical world (with objects, resources, etc.) and\nthe social world (with human demand, social behavior, etc.), and enhance smart\nresource allocation, automatic network operation, and intelligent service\nprovisioning.\n", " Revised formulas for the inclusive cross section of a triple parton\nscattering process in a hadron collision are suggested basing on the modified\ncollinear three-parton distributions. The possible phenomenological issues are\ndiscussed.\n", " We provide a probabilistic interpretation of attention and show that the\nstandard dot-product attention in transformers is a special case of Maximum A\nPosteriori (MAP) inference. The proposed approach suggests the use of\nExpectation Maximization algorithms for online adaptation of key and value\nmodel parameters. This approach is useful for cases in which external agents,\ne.g., annotators, provide inference-time information about the correct values\nof some tokens, e.g, the semantic category of some pixels, and we need for this\nnew information to propagate to other tokens in a principled manner. We\nillustrate the approach on an interactive semantic segmentation task in which\nannotators and models collaborate online to improve annotation efficiency.\nUsing standard benchmarks, we observe that key adaptation boosts model\nperformance ($\\sim10\\%$ mIoU) in the low feedback regime and value propagation\nimproves model responsiveness in the high feedback regime. A PyTorch layer\nimplementation of our probabilistic attention model will be made publicly\navailable here: https://github.com/apple/ml-probabilistic-attention.\n", " A real-world text corpus sometimes comprises not only text documents but also\nsemantic links between them (e.g., academic papers in a bibliographic network\nare linked by citations and co-authorships). Text documents and semantic\nconnections form a text-rich network, which empowers a wide range of downstream\ntasks such as classification and retrieval. However, pretraining methods for\nsuch structures are still lacking, making it difficult to build one generic\nmodel that can be adapted to various tasks on text-rich networks. Current\npretraining objectives, such as masked language modeling, purely model texts\nand do not take inter-document structure information into consideration. To\nthis end, we propose our PretrAining on TexT-Rich NetwOrk framework Patton.\nPatton includes two pretraining strategies: network-contextualized masked\nlanguage modeling and masked node prediction, to capture the inherent\ndependency between textual attributes and network structure. We conduct\nexperiments on four downstream tasks in five datasets from both academic and\ne-commerce domains, where Patton outperforms baselines significantly and\nconsistently.\n", " In this paper we show how inter-cellular molecular communication may change\nthe overall levels of photosynthesis in plants. Individual plant cells respond\nto external stimuli, such as illumination levels, to regulate their\nphotosynthetic output. Here, we present a mathematical model which shows that\nby sharing information internally using molecular communication, plants may\nincrease overall photosynthate production. Numerical results show that higher\nmutual information between cells corresponds to an increase in overall\nphotosynthesis by as much as 25 per cent. This suggests that molecular\ncommunication plays a vital role in maximising the photosynthesis in plants and\ntherefore suggests new routes to influence plant development in agriculture and\nelsewhere.\n", " Temporal action detection (TAD) involves the localization and classification\nof action instances within untrimmed videos. While standard TAD follows fully\nsupervised learning with closed-set setting on large training data, recent\nzero-shot TAD methods showcase the promising open-set setting by leveraging\nlarge-scale contrastive visual-language (ViL) pretrained models. However,\nexisting zero-shot TAD methods have limitations on how to properly construct\nthe strong relationship between two interdependent tasks of localization and\nclassification and adapt ViL model to video understanding. In this work, we\npresent ZEETAD, featuring two modules: dual-localization and zero-shot proposal\nclassification. The former is a Transformer-based module that detects action\nevents while selectively collecting crucial semantic embeddings for later\nrecognition. The latter one, CLIP-based module, generates semantic embeddings\nfrom text and frame inputs for each temporal unit. Additionally, we enhance\ndiscriminative capability on unseen classes by minimally updating the frozen\nCLIP encoder with lightweight adapters. Extensive experiments on THUMOS14 and\nActivityNet-1.3 datasets demonstrate our approach's superior performance in\nzero-shot TAD and effective knowledge transfer from ViL models to unseen action\ncategories.\n", " In the context of increasing demands for long-term multi-energy load\nforecasting in real-world applications, this paper introduces Patchformer, a\nnovel model that integrates patch embedding with encoder-decoder\nTransformer-based architectures. To address the limitation in existing\nTransformer-based models, which struggle with intricate temporal patterns in\nlong-term forecasting, Patchformer employs patch embedding, which predicts\nmultivariate time-series data by separating it into multiple univariate data\nand segmenting each of them into multiple patches. This method effectively\nenhances the model's ability to capture local and global semantic dependencies.\nThe numerical analysis shows that the Patchformer obtains overall better\nprediction accuracy in both multivariate and univariate long-term forecasting\non the novel Multi-Energy dataset and other benchmark datasets. In addition,\nthe positive effect of the interdependence among energy-related products on the\nperformance of long-term time-series forecasting across Patchformer and other\ncompared models is discovered, and the superiority of the Patchformer against\nother models is also demonstrated, which presents a significant advancement in\nhandling the interdependence and complexities of long-term multi-energy\nforecasting. Lastly, Patchformer is illustrated as the only model that follows\nthe positive correlation between model performance and the length of the past\nsequence, which states its ability to capture long-range past local semantic\ninformation.\n", " The effect of signal integration through cis-regulatory modules (CRMs) on\nsynchronization and clustering of populations of two-component genetic\noscillators coupled by quorum sensing is in detail investigated. We find that\nthe CRMs play an important role in achieving synchronization and clustering.\nFor this, we investigate 6 possible cis-regulatory input functions (CRIFs) with\nAND, OR, ANDN, ORN, XOR, and EQU types of responses in two possible kinds of\ncell-to-cell communications: activator-regulated communication (i.e., the\nautoinducer regulates the activator) and repressor-regulated communication\n(i.e., the autoinducer regulates the repressor). Both theoretical analysis and\nnumerical simulation show that different CRMs drive fundamentally different\ncellular patterns, such as complete synchronization, various cluster-balanced\nstates and several cluster-nonbalanced states.\n", " Humans can reason compositionally when presented with new tasks. Previous\nresearch shows that appropriate prompting techniques enable large language\nmodels (LLMs) to solve artificial compositional generalization tasks such as\nSCAN. In this work, we identify additional challenges in more realistic\nsemantic parsing tasks with larger vocabulary and refine these prompting\ntechniques to address them. Our best method is based on least-to-most\nprompting: it decomposes the problem using prompting-based syntactic parsing,\nthen uses this decomposition to select appropriate exemplars and to\nsequentially generate the semantic parse. This method allows us to set a new\nstate of the art for CFQ while requiring only 1% of the training data used by\ntraditional approaches. Due to the general nature of our approach, we expect\nsimilar efforts will lead to new results in other tasks and domains, especially\nfor knowledge-intensive applications.\n", " I review progress on selected issues connected with generalized parton\ndistributions. Topics range from the description of hard exclusive reactions to\nthe spatial distribution of quarks in the nucleon and the contribution of their\norbital angular momentum to the nucleon spin.\n", " Recent advances in Vision Transformers (ViT) and Stable Diffusion (SD) models\nwith their ability to capture rich semantic features of the image have been\nused for image correspondence tasks on natural images. In this paper, we\nexamine the ability of a variety of pre-trained ViT (DINO, DINOv2, SAM, CLIP)\nand SD models, trained exclusively on natural images, for solving the\ncorrespondence problems on medical images. While many works have made a case\nfor in-domain training, we show that the models trained on natural images can\noffer good performance on medical images across different modalities\n(CT,MR,Ultrasound) sourced from various manufacturers, over multiple anatomical\nregions (brain, thorax, abdomen, extremities), and on wide variety of tasks.\nFurther, we leverage the correspondence with respect to a template image to\nprompt a Segment Anything (SAM) model to arrive at single shot segmentation,\nachieving dice range of 62%-90% across tasks, using just one image as\nreference. We also show that our single-shot method outperforms the recently\nproposed few-shot segmentation method - UniverSeg (Dice range 47%-80%) on most\nof the semantic segmentation tasks(six out of seven) across medical imaging\nmodalities.\n", " The Wiener index of a network, introduced by the chemist Harry Wiener, is the\nsum of distances between all pairs of nodes in the network. This index,\noriginally used in chemical graph representations of the non-hydrogen atoms of\na molecule, is considered to be a fundamental and useful network descriptor. We\nstudy the problem of constructing geometric networks on point sets in Euclidean\nspace that minimize the Wiener index: given a set $P$ of $n$ points in\n$\\mathbb{R}^d$, the goal is to construct a network, spanning $P$ and satisfying\ncertain constraints, that minimizes the Wiener index among the allowable class\nof spanning networks.\n In this work, we focus mainly on spanning networks that are trees and we\nfocus on problems in the plane ($d=2$). We show that any spanning tree that\nminimizes the Wiener index has non-crossing edges in the plane. Then, we use\nthis fact to devise an $O(n^4)$-time algorithm that constructs a spanning tree\nof minimum Wiener index for points in convex position. We also prove that the\nproblem of computing a spanning tree on $P$ whose Wiener index is at most $W$,\nwhile having total (Euclidean) weight at most $B$, is NP-hard.\n Computing a tree that minimizes the Wiener index has been studied in the area\nof communication networks, where it is known as the optimum communication\nspanning tree problem.\n", " Examining textbooks of thermodynamics we have not able to find a proof of\nincreasing of entropy under free adiabatic expansion in vacuum based on the\nstate equation of the Van der Waals gas. This communication remedies the\nsituation. During proving we faced with an amusing example of apparent\nviolation of thermodynamics second law.\n", " Advanced persistent threats (APTs) have novel features such as multi-stage\npenetration, highly-tailored intention, and evasive tactics. APTs defense\nrequires fusing multi-dimensional Cyber threat intelligence data to identify\nattack intentions and conducts efficient knowledge discovery strategies by\ndata-driven machine learning to recognize entity relationships. However,\ndata-driven machine learning lacks generalization ability on fresh or unknown\nsamples, reducing the accuracy and practicality of the defense model. Besides,\nthe private deployment of these APT defense models on heterogeneous\nenvironments and various network devices requires significant investment in\ncontext awareness (such as known attack entities, continuous network states,\nand current security strategies). In this paper, we propose a few-shot\nmulti-domain knowledge rearming (FMKR) scheme for context-aware defense against\nAPTs. By completing multiple small tasks that are generated from different\nnetwork domains with meta-learning, the FMKR firstly trains a model with good\ndiscrimination and generalization ability for fresh and unknown APT attacks. In\neach FMKR task, both threat intelligence and local entities are fused into the\nsupport/query sets in meta-learning to identify possible attack stages.\nSecondly, to rearm current security strategies, an finetuning-based deployment\nmechanism is proposed to transfer learned knowledge into the student model,\nwhile minimizing the defense cost. Compared to multiple model replacement\nstrategies, the FMKR provides a faster response to attack behaviors while\nconsuming less scheduling cost. Based on the feedback from multiple real users\nof the Industrial Internet of Things (IIoT) over 2 months, we demonstrate that\nthe proposed scheme can improve the defense satisfaction rate.\n", " Motivation: In transcriptomic and proteomics-based studies, the abundance of\ngenes is often compared to functional pathways such as the Kyoto Encyclopaedia\nat Genes and Genomes (KEGG) to identify active metabolic processes. Even though\na plethora of tools allow to analyze and to compare omics data in respect to\nKEGG pathways, the analysis of multiple conditions is often limited to only a\ndefined set of conditions. Furthermore, for transcriptomic datasets, it is\ncrucial to compare the entire set of pathways in order to obtain a global\noverview of the species' metabolic functions. Results: Here, we present the\ntool KEGGexpressionMapper, a module, that is implemented in the programming\nlanguage R. The module allows to highlight the expression of transcriptomic or\nproteomic measurements in various conditions on pathways and incorporates\nmethods to analyze gene enrichment analyses and expression clustering in time\nseries data. KEGGexpressionMapper supports time series data from transcriptomic\nor proteomics measurements from different individuals. As the tool is\nimplemented in the scripting language R, it can be integrated into existing\nanalysis pipelines to obtain a global overview of the dataset. The R package\ncan be downloaded from https://github.com/nthomasCUBE/KEGGexpressionmapper.\n", " We consider a mobile edge computing scenario where users want to perform a\nlinear inference operation $\\boldsymbol{W} \\boldsymbol{x}$ on local data\n$\\boldsymbol{x}$ for some network-side matrix $\\boldsymbol{W}$. The inference\nis performed in a distributed fashion over multiple servers at the network\nedge. For this scenario, we propose a coding scheme that combines a rateless\ncode to provide resiliency against straggling servers--hence reducing the\ncomputation latency--and an irregular-repetition code to provide spatial\ndiversity--hence reducing the communication latency. We further derive a lower\nbound on the total latency--comprising computation latency, communication\nlatency, and decoding latency. The proposed scheme performs remarkably close to\nthe bound and yields significantly lower latency than the scheme based on\nmaximum distance separable codes recently proposed by Zhang and Simeone.\n", " Gillevinia straata, the scientific name [1, 2] recognizing the first\nextraterrestrial living form ever nomenclated, as well as the existence of a\nnew biological kingdom, Jakobia, in a new biosphere -Marciana- of what now has\nbecome the living system Solaria, is grounded on old evidence reinterpreted in\nthe light of newly acquired facts. The present exposition provides a summary\noverview of all these grounds, outlined here as follows. A more detailed paper\nis being prepared for publication.\n", " We continue a study of unconditionally secure all-or-nothing transforms\n(AONT) begun in \\cite{St}. An AONT is a bijective mapping that constructs s\noutputs from s inputs. We consider the security of t inputs, when s-t outputs\nare known. Previous work concerned the case t=1; here we consider the problem\nfor general t, focussing on the case t=2. We investigate constructions of\nbinary matrices for which the desired properties hold with the maximum\nprobability. Upper bounds on these probabilities are obtained via a quadratic\nprogramming approach, while lower bounds can be obtained from combinatorial\nconstructions based on symmetric BIBDs and cyclotomy. We also report some\nresults on exhaustive searches and random constructions for small values of s.\n", " Prominent AI experts have suggested that companies developing high-risk AI\nsystems should be required to show that such systems are safe before they can\nbe developed or deployed. The goal of this paper is to expand on this idea and\nexplore its implications for risk management. We argue that entities developing\nor deploying high-risk AI systems should be required to present evidence of\naffirmative safety: a proactive case that their activities keep risks below\nacceptable thresholds. We begin the paper by highlighting global security risks\nfrom AI that have been acknowledged by AI experts and world governments. Next,\nwe briefly describe principles of risk management from other high-risk fields\n(e.g., nuclear safety). Then, we propose a risk management approach for\nadvanced AI in which model developers must provide evidence that their\nactivities keep certain risks below regulator-set thresholds. As a first step\ntoward understanding what affirmative safety cases should include, we\nillustrate how certain kinds of technical evidence and operational evidence can\nsupport an affirmative safety case. In the technical section, we discuss\nbehavioral evidence (evidence about model outputs), cognitive evidence\n(evidence about model internals), and developmental evidence (evidence about\nthe training process). In the operational section, we offer examples of\norganizational practices that could contribute to affirmative safety cases:\ninformation security practices, safety culture, and emergency response\ncapacity. Finally, we briefly compare our approach to the NIST AI Risk\nManagement Framework. Overall, we hope our work contributes to ongoing\ndiscussions about national and global security risks posed by AI and regulatory\napproaches to address these risks.\n", "4z(Assadi, Chen, and Khanna, 2019) define a 4-player hidden-pointer-chasing\n($\\mathsf{HPC}^4$), and using it, give strong multi-pass lower bounds for graph\nproblems in the streaming model of computation and a lower bound on the query\ncomplexity of sub-modular minimization. We present a two-player version\n($\\mathsf{HPC}^2$) of $\\mathsf{HPC}^4$ that has matching communication\ncomplexity to $\\mathsf{HPC}^4$. Our formulation allows us to lower bound its\ncommunication complexity with a simple direct-sum argument. Using this lower\nbound on the communication complexity of $\\mathsf{HPC}^2$, we retain the\nstreaming and query complexity lower bounds by (Assadi, Chen, and Khanna,\n2019).\n Further, by giving reductions from $\\mathsf{HPC}^2$, we prove new multi-pass\nspace lower bounds for graph problems in turnstile streams. In particular, we\nshow that any algorithm which computes the exact weight of the maximum weighted\nmatching in an $n$-vertex graph requires $\\tilde{O}(n^{2})$ space unless it\nmakes $\\omega(\\log n)$ passes over the turnstile stream, and that any algorithm\nwhich computes the minimum $s\\text{-}t$ distance in an $n$-vertex graph\nrequires $n^{2-o(1)}$ space unless it makes $n^{\\Omega(1)}$ passes over the\nturnstile stream. Our reductions can be modified to use $\\mathsf{HPC}^4$ as\nwell.\n", " We show that quantum communication by means of collapse of the wave function\nis possible. In this study, quantum communication does not mean quantum\nteleportation or quantum cryptography, but transmission of information itself.\nBecause of consistency with special relativity, the possibility of the quantum\ncommunication leads to another conclusion that the collapse of the wave\nfunction must propagate at the speed of light or slower.\n We show this requirement is consistent with nonlocality in quantum mechanics.\nWe also demonstrate that the Einstein-Podolsky-Rosen experiment does not\ndisprove our conclusion.\n", " Relationships are essential to our happiness and wellbeing. The dissolution\nof a relationship, the final stage of relationship's lifecycle and one of the\nmost stressful events in an individual's life, can have profound and\nlong-lasting impacts on people. With the breakup process increasingly\nfacilitated by computer-mediated communication (CMC), and the likely future\ninfluence of AI-mediated communication (AIMC) tools, we conducted a\nsemi-structured interview study with 21 participants. We aim to understand: 1)\nthe current role of technology in the breakup process, 2) the needs and support\nindividuals have during the process, and 3) how AI might address these needs.\nOur research shows that people have distinct needs at various stages of ending\na relationship. Presently, technology is used for information gathering and\ncommunity support, acting as a catalyst for breakups, enabling ghosting and\nblocking, and facilitating communication. Participants anticipate that AI could\naid in sense-making of their relationship leading up to the breakup, act as a\nmediator, assist in crafting appropriate wording, tones, and language during\nbreakup conversations, and support companionship, reflection, recovery, and\ngrowth after a breakup. Our findings also demonstrate an overlap between the\nbreakup process and the Transtheoretical Model (TTM) of behavior change.\nThrough the lens of TTM, we explore the potential support and affordances AI\ncould offer in breakups, including its benefits and the necessary precautions\nregarding AI's role in this sensitive process.\n", " Diffusion models have demonstrated excellent performance in image generation.\nAlthough various few-shot semantic segmentation (FSS) models with different\nnetwork structures have been proposed, performance improvement has reached a\nbottleneck. This paper presents the first work to leverage the diffusion model\nfor FSS task, called DifFSS. DifFSS, a novel FSS paradigm, can further improve\nthe performance of the state-of-the-art FSS models by a large margin without\nmodifying their network structure. Specifically, we utilize the powerful\ngeneration ability of diffusion models to generate diverse auxiliary support\nimages by using the semantic mask, scribble or soft HED boundary of the support\nimage as control conditions. This generation process simulates the variety\nwithin the class of the query image, such as color, texture variation,\nlighting, $etc$. As a result, FSS models can refer to more diverse support\nimages, yielding more robust representations, thereby achieving a consistent\nimprovement in segmentation performance. Extensive experiments on three\npublicly available datasets based on existing advanced FSS models demonstrate\nthe effectiveness of the diffusion model for FSS task. Furthermore, we explore\nin detail the impact of different input settings of the diffusion model on\nsegmentation performance. Hopefully, this completely new paradigm will bring\ninspiration to the study of FSS task integrated with AI-generated content. Code\nis available at https://github.com/TrinitialChan/DifFSS\n", " In the wake of the burgeoning expansion of generative artificial intelligence\n(AI) services, the computational demands inherent to these technologies\nfrequently necessitate cloud-powered computational offloading, particularly for\nresource-constrained mobile devices. These services commonly employ prompts to\nsteer the generative process, and both the prompts and the resultant content,\nsuch as text and images, may harbor privacy-sensitive or confidential\ninformation, thereby elevating security and privacy risks. To mitigate these\nconcerns, we introduce $\\Lambda$-Split, a split computing framework to\nfacilitate computational offloading while simultaneously fortifying data\nprivacy against risks such as eavesdropping and unauthorized access. In\n$\\Lambda$-Split, a generative model, usually a deep neural network (DNN), is\npartitioned into three sub-models and distributed across the user's local\ndevice and a cloud server: the input-side and output-side sub-models are\nallocated to the local, while the intermediate, computationally-intensive\nsub-model resides on the cloud server. This architecture ensures that only the\nhidden layer outputs are transmitted, thereby preventing the external\ntransmission of privacy-sensitive raw input and output data. Given the\nblack-box nature of DNNs, estimating the original input or output from\nintercepted hidden layer outputs poses a significant challenge for malicious\neavesdroppers. Moreover, $\\Lambda$-Split is orthogonal to traditional\nencryption-based security mechanisms, offering enhanced security when deployed\nin conjunction. We empirically validate the efficacy of the $\\Lambda$-Split\nframework using Llama 2 and Stable Diffusion XL, representative large language\nand diffusion models developed by Meta and Stability AI, respectively. Our\n$\\Lambda$-Split implementation is publicly accessible at\nhttps://github.com/nishio-laboratory/lambda_split.\n", " We investigate the secure connectivity of wireless sensor networks under the\npairwise key distribution scheme of Chan et al.. Unlike recent work which was\ncarried out under the assumption of full visibility, here we assume a\n(simplified) communication model where unreliable wireless links are\nrepresented as on/off channels. We present conditions on how to scale the model\nparameters so that the network i) has no secure node which is isolated and ii)\nis securely connected, both with high probability when the number of sensor\nnodes becomes large. The results are given in the form of zero-one laws, and\nexhibit significant differences with corresponding results in the full\nvisibility case. Through simulations these zero-one laws are shown to be valid\nalso under a more realistic communication model, i.e., the disk model.\n", " Meaning has been called the \"holy grail\" of a variety of scientific\ndisciplines, ranging from linguistics to philosophy, psychology and the\nneurosciences. The field of Artifical Intelligence (AI) is very much a part of\nthat list: the development of sophisticated natural language semantics is a\nsine qua non for achieving a level of intelligence comparable to humans.\nEmbodiment theories in cognitive science hold that human semantic\nrepresentation depends on sensori-motor experience; the abundant evidence that\nhuman meaning representation is grounded in the perception of physical reality\nleads to the conclusion that meaning must depend on a fusion of multiple\n(perceptual) modalities. Despite this, AI research in general, and its\nsubdisciplines such as computational linguistics and computer vision in\nparticular, have focused primarily on tasks that involve a single modality.\nHere, we propose virtual embodiment as an alternative, long-term strategy for\nAI research that is multi-modal in nature and that allows for the kind of\nscalability required to develop the field coherently and incrementally, in an\nethically responsible fashion.\n", " Vertical federated learning (VFL), where data features are stored in multiple\nparties distributively, is an important area in machine learning. However, the\ncommunication complexity for VFL is typically very high. In this paper, we\npropose a unified framework by constructing coresets in a distributed fashion\nfor communication-efficient VFL. We study two important learning tasks in the\nVFL setting: regularized linear regression and $k$-means clustering, and apply\nour coreset framework to both problems. We theoretically show that using\ncoresets can drastically alleviate the communication complexity, while nearly\nmaintain the solution quality. Numerical experiments are conducted to\ncorroborate our theoretical findings.\n", " We study dynamic interplay between time-delay and velocity alignment in the\nensemble of Cucker-Smale (C-S) particles(or agents) on time-varying networks\nwhich are modeled by digraphs containing spanning trees. Time-delayed dynamical\nsystems often appear in mathematical models from biology and control theory,\nand they have been extensively investigated in literature. In this paper, we\nprovide sufficient frameworks for the mono-cluster flocking to the continuous\nand discrete C-S models, which are formulated in terms of system parameters and\ninitial data. In our proposed frameworks, we show that the continuous and\ndiscrete C-S models exhibit exponential flocking estimates. For the explicit\nC-S communication weights which decay algebraically, our results exhibit\nthreshold phenomena depending on the decay rate and depth of digraph. We also\nprovide several numerical examples and compare them with our analytical\nresults.\n", " We study multi-modal few-shot object detection (FSOD) in this paper, using\nboth few-shot visual examples and class semantic information for detection,\nwhich are complementary to each other by definition. Most of the previous works\non multi-modal FSOD are fine-tuning-based which are inefficient for online\napplications. Moreover, these methods usually require expertise like class\nnames to extract class semantic embedding, which are hard to get for rare\nclasses. Our approach is motivated by the high-level conceptual similarity of\n(metric-based) meta-learning and prompt-based learning to learn generalizable\nfew-shot and zero-shot object detection models respectively without\nfine-tuning. Specifically, we combine the few-shot visual classifier and text\nclassifier learned via meta-learning and prompt-based learning respectively to\nbuild the multi-modal classifier and detection models. In addition, to fully\nexploit the pre-trained language models, we propose meta-learning-based\ncross-modal prompting to generate soft prompts for novel classes present in\nfew-shot visual examples, which are then used to learn the text classifier.\nKnowledge distillation is introduced to learn the soft prompt generator without\nusing human prior knowledge of class names, which may not be available for rare\nclasses. Our insight is that the few-shot support images naturally include\nrelated context information and semantics of the class. We comprehensively\nevaluate the proposed multi-modal FSOD models on multiple few-shot object\ndetection benchmarks, achieving promising results.\n", " Remarkable effectiveness of the channel or spatial attention mechanisms for\nproducing more discernible feature representation are illustrated in various\ncomputer vision tasks. However, modeling the cross-channel relationships with\nchannel dimensionality reduction may bring side effect in extracting deep\nvisual representations. In this paper, a novel efficient multi-scale attention\n(EMA) module is proposed. Focusing on retaining the information on per channel\nand decreasing the computational overhead, we reshape the partly channels into\nthe batch dimensions and group the channel dimensions into multiple\nsub-features which make the spatial semantic features well-distributed inside\neach feature group. Specifically, apart from encoding the global information to\nre-calibrate the channel-wise weight in each parallel branch, the output\nfeatures of the two parallel branches are further aggregated by a\ncross-dimension interaction for capturing pixel-level pairwise relationship. We\nconduct extensive ablation studies and experiments on image classification and\nobject detection tasks with popular benchmarks (e.g., CIFAR-100, ImageNet-1k,\nMS COCO and VisDrone2019) for evaluating its performance.\n", " Some precisions are given about the definition of the Hamiltonian operator H\nand its transformation properties, for a linear wave equation in a general\nspacetime. In the presence of time-dependent unitary gauge transformations, H\nas an operator depends on the gauge choice. The other observables of QM and\ntheir rates also become gauge-dependent unless a proper account for the gauge\nchoice is done in their definition. We show the explicit effect of these\nnon-uniqueness issues in the case of the Dirac equation in a general spacetime\nwith the Schwinger gauge. We show also in detail why, the meaning of the energy\nin QM being inherited from classical Hamiltonian mechanics, the energy operator\nand its mean values ought to be well defined in a general spacetime.\n", " Federated learning (FL) brings collaborative intelligence into industries\nwithout centralized training data to accelerate the process of Industry 4.0 on\nthe edge computing level. FL solves the dilemma in which enterprises wish to\nmake the use of data intelligence with security concerns. To accelerate\nindustrial Internet of things with the further leverage of FL, existing\nachievements on FL are developed from three aspects: 1) define terminologies\nand elaborate a general framework of FL for accommodating various scenarios; 2)\ndiscuss the state-of-the-art of FL on fundamental researches including data\npartitioning, privacy preservation, model optimization, local model\ntransportation, personalization, motivation mechanism, platform & tools, and\nbenchmark; 3) discuss the impacts of FL from the economic perspective. To\nattract more attention from industrial academia and practice, a FL-transformed\nmanufacturing paradigm is presented, and future research directions of FL are\ngiven and possible immediate applications in Industry 4.0 domain are also\nproposed.\n", " An overview is given on key physics, detector and accelerator aspects of the\nLHeC, including its further development, with emphasis to its role as the\ncleanest microscope of parton dynamics and a precision Higgs facility.\n", " Effective field theory (EFT) formulations of dark matter interactions have\nproven to be a convenient and popular way to quantify LHC bounds on dark\nmatter. However, some of the non-renormalizable EFT operators considered do not\nrespect the gauge symmetries of the Standard Model. We carefully discuss under\nwhat circumstances such operators can arise, and outline potential issues in\ntheir interpretation and application.\n", " In wireless networks, energy consumed for communication includes both the\ntransmission and the processing energy. In this paper, point-to-point\ncommunication over a fading channel with an energy harvesting transmitter is\nstudied considering jointly the energy costs of transmission and processing.\nUnder the assumption of known energy arrival and fading profiles, optimal\ntransmission policy for throughput maximization is investigated. Assuming that\nthe transmitter has sufficient amount of data in its buffer at the beginning of\nthe transmission period, the average throughput by a given deadline is\nmaximized. Furthermore, a \"directional glue pouring algorithm\" that computes\nthe optimal transmission policy is described.\n" ]

[ " The gains afforded by cloud radio access network (C-RAN) in terms of savings\nin capital and operating expenses, flexibility, interference management and\nnetwork densification rely on the presence of high-capacity low-latency\nfronthaul connectivity between remote radio heads (RRHs) and baseband unit\n(BBU). In light of the non-uniform and limited availability of fiber optics\ncables, the bandwidth constraints on the fronthaul network call, on the one\nhand, for the development of advanced baseband compression strategies and, on\nthe other hand, for a closer investigation of the optimal functional split\nbetween RRHs and BBU. In this chapter, after a brief introduction to signal\nprocessing challenges in C-RAN, this optimal function split is studied at the\nphysical (PHY) layer as it pertains to two key baseband signal processing\nsteps, namely channel estimation in the uplink and channel encoding/ linear\nprecoding in the downlink. Joint optimization of baseband fronthaul compression\nand of baseband signal processing is tackled under different PHY functional\nsplits, whereby uplink channel estimation and downlink channel encoding/ linear\nprecoding are carried out either at the RRHs or at the BBU. The analysis, based\non information-theoretical arguments, and numerical results yields insight into\nthe configurations of network architecture and fronthaul capacities in which\ndifferent functional splits are advantageous. The treatment also emphasizes the\nversatility of deterministic and stochastic successive convex approximation\nstrategies for the optimization of C-RANs.\n", " In some estimation problems, not all the parameters can be identified, which\nresults in singularity of the Fisher Information Matrix (FIM). The Cram\\'er-Rao\nBound (CRB), which is the inverse of the FIM, is then not defined. To\nregularize the estimation problem, one can impose constraints on the parameters\nand derive the corresponding CRBs. The correspondence between local\nidentifiability and FIM regularity is studied here. Furthermore the number of\nFIM singularities is shown to be equal to the number of independent constraints\nnecessary to have a well-defined constrained CRB and local identifiability. In\ngeneral, many sets of constraints can render the parameters identifiable,\ngiving different values for the CRB, that are not always relevant. When the\nconstraints can be chosen, we propose a constrained CRB, the pseudo-inverse of\nthe FIM, which gives, for a minimum number of constraints, the lowest bound on\nthe mean squared estimation error. These results are applied to two approaches\nto blind FIR multichannel estimation which allow identification of the channel\nup to a scale or phase factor. These two approaches correspond to deterministic\nand Gaussian models for the unknown channel inputs. The singularities of the\nFIMs and local identifiability are studied and the corresponding constrained\nCRBs are derived and interpreted.\n", " Recent interest in integrated sensing and communications has led to the\ndesign of novel signal processing techniques to recover information from an\noverlaid radar-communications signal. Here, we focus on a spectral coexistence\nscenario, wherein the channels and transmit signals of both radar and\ncommunications systems are unknown to the common receiver. In this dual-blind\ndeconvolution (DBD) problem, the receiver admits a multi-carrier wireless\ncommunications signal that is overlaid with the radar signal reflected off\nmultiple targets. The communications and radar channels are represented by\ncontinuous-valued range-times or delays corresponding to multiple transmission\npaths and targets, respectively. Prior works addressed recovery of unknown\nchannels and signals in this ill-posed DBD problem through atomic norm\nminimization but contingent on individual minimum separation conditions for\nradar and communications channels. In this paper, we provide an optimal joint\nseparation condition using extremal functions from the Beurling-Selberg\ninterpolation theory. Thereafter, we formulate DBD as a low-rank modified\nHankel matrix retrieval and solve it via nuclear norm minimization. We estimate\nthe unknown target and communications parameters from the recovered low-rank\nmatrix using multiple signal classification (MUSIC) method. We show that the\njoint separation condition also guarantees that the underlying Vandermonde\nmatrix for MUSIC is well-conditioned. Numerical experiments validate our\ntheoretical findings.\n", " Semantic communication allows the receiver to know the intention instead of\nthe bit information itself, which is an emerging technique to support real-time\nhuman-machine and machine-to-machine interactions for future wireless\ncommunications. In semantic communications, both transmitter and receiver share\nsome common knowledge, which can be used to extract small-size information at\nthe transmitter and recover the original information at the receiver. Due to\ndifferent design purposes, security issues in semantic communications have two\nunique features compared to standard bit-wise communications. First, an\nattacker in semantic communications considers not only the amount of stolen\ndata but also the meanings of stolen data. Second, an attacker in semantic\ncommunication systems can attack not only semantic information transmission as\ndone in standard communication systems but also attacks machine learning (ML)\nmodels used for semantic information extraction since most of semantic\ninformation is generated using ML based methods. Due to these unique features,\nin this paper, we present an overview on the fundamentals and key challenges in\nthe design of secure semantic communication. We first provide various methods\nto define and extract semantic information. Then, we focus on secure semantic\ncommunication techniques in two areas: information security and semantic ML\nmodel security. For each area, we identify the main problems and challenges.\nThen, we will provide a comprehensive treatment of these problems. In a\nnutshell,this article provides a holistic set of guidelines on how to design\nsecure semantic communication systems over real-world wireless communication\nnetworks.\n", " Audio applications involving environmental sound analysis increasingly use\ngeneral-purpose audio representations, also known as embeddings, for transfer\nlearning. Recently, Holistic Evaluation of Audio Representations (HEAR)\nevaluated twenty-nine embedding models on nineteen diverse tasks. However, the\nevaluation's effectiveness depends on the variation already captured within a\ngiven dataset. Therefore, for a given data domain, it is unclear how the\nrepresentations would be affected by the variations caused by myriad\nmicrophones' range and acoustic conditions -- commonly known as channel\neffects. We aim to extend HEAR to evaluate invariance to channel effects in\nthis work. To accomplish this, we imitate channel effects by injecting\nperturbations to the audio signal and measure the shift in the new (perturbed)\nembeddings with three distance measures, making the evaluation domain-dependent\nbut not task-dependent. Combined with the downstream performance, it helps us\nmake a more informed prediction of how robust the embeddings are to the channel\neffects. We evaluate two embeddings -- YAMNet, and OpenL3 on monophonic\n(UrbanSound8K) and polyphonic (SONYC-UST) urban datasets. We show that one\ndistance measure does not suffice in such task-independent evaluation. Although\nFr\\'echet Audio Distance (FAD) correlates with the trend of the performance\ndrop in the downstream task most accurately, we show that we need to study FAD\nin conjunction with the other distances to get a clear understanding of the\noverall effect of the perturbation. In terms of the embedding performance, we\nfind OpenL3 to be more robust than YAMNet, which aligns with the HEAR\nevaluation.\n", " We prove an affine analog of Scharlau's reduction theorem for spherical\nbuildings. To be a bit more precise let $X$ be a euclidean building with\nspherical building $\\partial X$ at infinity. Then there exists a euclidean\nbuilding $\\bar X$ such that $X$ splits as a product of $\\bar X$ with some\neuclidean $k$-space such that $\\partial \\bar X$ is the thick reduction of\n$\\partial X$ in the sense of Scharlau. \\newline In addition we prove a converse\nstatement saying that an embedding of a thick spherical building at infinity\nextends to an embedding of the euclidean building having the extended spherical\nbuilding as its boundary.\n", " One of the first widespread uses of multi-user multiple-input multiple-output\n(MU-MIMO) is in 5G networks, where each base station has an advanced antenna\nsystem (AAS) that is connected to the baseband unit (BBU) with a\ncapacity-constrained fronthaul. In the AAS configuration, multiple passive\nantenna elements and radio units are integrated into a single box. This paper\nconsiders precoded downlink transmission over a single-cell MU-MIMO system. We\nstudy optimized linear precoding for AAS with a limited-capacity fronthaul,\nwhich requires the precoding matrix to be quantized. We propose a new precoding\ndesign that is aware of the fronthaul quantization and minimizes the\nmean-squared error at the receiver side. We compute the precoding matrix using\na sphere decoding (SD) approach. We also propose a heuristic low-complexity\napproach to quantized precoding. This heuristic is computationally efficient\nenough for massive MIMO systems. The numerical results show that our proposed\nprecoding significantly outperforms quantization-unaware precoding and other\nprevious approaches in terms of the sum rate. The performance loss for our\nheuristic method compared to quantization-aware precoding is insignificant\nconsidering the complexity reduction, which makes the heuristic method feasible\nfor real-time applications. We consider both perfect and imperfect channel\nstate information.\n", " With ever increasing data rates produced by modern radio telescopes like\nLOFAR and future telescopes like the SKA, many data processing steps are\noverwhelmed by the amount of data that needs to be handled using limited\ncompute resources. Calibration is one such operation that dominates the overall\ndata processing computational cost, nonetheless, it is an essential operation\nto reach many science goals. Calibration algorithms do exist that scale well\nwith the number of stations of an array and the number of directions being\ncalibrated. However, the remaining bottleneck is the raw data volume, which\nscales with the number of baselines, and which is proportional to the square of\nthe number of stations. We propose a 'stochastic' calibration strategy where we\nonly read in a mini-batch of data for obtaining calibration solutions, as\nopposed to reading the full batch of data being calibrated. Nonetheless, we\nobtain solutions that are valid for the full batch of data. Normally, data need\nto be averaged before calibration is performed to accommodate the data in\nsize-limited compute memory. Stochastic calibration overcomes the need for data\naveraging before any calibration can be performed, and offers many advantages\nincluding: enabling the mitigation of faint radio frequency interference;\nbetter removal of strong celestial sources from the data; and better detection\nand spatial localization of fast radio transients.\n", " In this paper, forward and backward propagating waves and reflectivity in an\noptical waveguide structure namely the fiber Bragg reflector also considered as\na one dimensional photonic crystal, are analytically computed using coupled\nmode theory for different grating lengths and coupling conditions.\nAlxGa1-xN/GaN material composition is considered as unit block of the periodic\norganization, and refractive index of AlxGa1-xN is taken to be dependent on\nmaterial composition, bandgap and operating wavelength following Adachis'\nmodel. The structure being considered is the Bragg grating where increase in\ngrating length enhances the reflection of electromagnetic wave, and strong\ncoupling provides larger bandgap spectral width. Input wavelength is made\ndifferent from Bragg wavelength to study the characteristics of propagating\nwaves. A suitable combination of grating length and coupling coefficient is\nhelpful in designing the photonic bandgap at 1550 nm wavelength. These\ncharacteristic curves can be utilized to study how waves propagate through the\noptical waveguides which have a special place in optical communications\n", " Mixing (or quasirandom) properties of the natural transition matrix\nassociated to a graph can be quantified by its distance to the complete graph.\nDifferent mixing properties correspond to different norms to measure this\ndistance. For dense graphs, two such properties known as spectral expansion and\nuniformity were shown to be equivalent in seminal 1989 work of Chung, Graham\nand Wilson. Recently, Conlon and Zhao extended this equivalence to the case of\nsparse vertex transitive graphs using the famous Grothendieck inequality. Here\nwe generalize these results to the non-commutative, or `quantum', case, where a\ntransition matrix becomes a quantum channel. In particular, we show that for\nirreducibly covariant quantum channels, expansion is equivalent to a natural\nanalog of uniformity for graphs, generalizing the result of Conlon and Zhao.\nMoreover, we show that in these results, the non-commutative and commutative\n(resp.) Grothendieck inequalities yield the best-possible constants.\n", " We model biochemical signal transduction, based on a ligand-receptor binding\nmechanism, as a discrete-time finite-state Markov channel, which we call the\nBIND channel. We show how to obtain the capacity of this channel, for the case\nof binary output, binary channel state, and arbitrary finite input alphabets.\nWe show that the capacity-achieving input distribution is IID. Further, we show\nthat feedback does not increase the capacity of this channel. We show how the\ncapacity of the discrete-time channel approaches the capacity of Kabanov's\nPoisson channel, in the limit of short time steps and rapid ligand release.\n", " A new approach for the approximation of the channel log-likelihood ratio\n(LLR) for wireless channels based on Taylor series is proposed. The\napproximation is applied to the uncorrelated flat Rayleigh fading channel with\nunknown channel state information at the receiver. It is shown that the\nproposed approximation greatly simplifies the calculation of channel LLRs, and\nyet provides results almost identical to those based on the exact calculation\nof channel LLRs. The results are obtained in the context of bit-interleaved\ncoded modulation (BICM) schemes with low-density parity-check (LDPC) codes, and\ninclude threshold calculations and error rate performance of finite-length\ncodes. Compared to the existing approximations, the proposed method is either\nsignificantly less complex, or considerably more accurate.\n", " We consider chemical reaction networks modeled by a discrete state and\ncontinuous in time Markov process for the vector copy number of the species and\nprovide a novel particle filter method for state and parameter estimation based\non exact observation of some of the species in continuous time. The conditional\nprobability distribution of the unobserved states is shown to satisfy a system\nof differential equations with jumps. We provide a method of simulating a\nprocess that is a proxy for the vector copy number of the unobserved species\nalong with a weight. The resulting weighted Monte Carlo simulation is then used\nto compute the conditional probability distribution of the unobserved species.\nWe also show how our algorithm can be adapted for a Bayesian estimation of\nparameters and for the estimation of a past state value based on observations\nup to a future time.\n", " Single photon emitters with narrow linewidths are highly sought after for\napplications in quantum information processing and quantum communications. In\nthis letter, we report on a bright, highly polarized near infrared single\nphoton emitter embedded in diamond nanocrystals with a narrow, sub GHz optical\nlinewidths at 10K. The observed zero phonon line at ~ 780 nm is optically\nstable under low power resonant excitation and blue shifts as the excitation\npower increases. Our results highlight the prospect for using new near infrared\ncolor centers in nanodiamonds for quantum applications.\n", " The Darlington synthesis problem (in the scalar case) is the problem of\nembedding a given contractive analytic function to an inner $2\\times 2$ matrix\nfunction as the entry. A fundamental result of Arov--Douglas--Helton relates\nthis algebraic property to a pure analytic one known as a pseudocontinuation of\nbounded type. We suggest a local version of the Darlington synthesis problem\nand prove a local analog of the ADH theorem.\n", " The main result of these notes is an analytical expression for the partition\nfunction of the circular brane model for arbitrary values of the topological\nangle. The model has important applications in condensed matter physics. It is\nrelated to the dissipative rotator (Ambegaokar-Eckern-Schon) model and\ndescribes a ``weakly blocked'' quantum dot with an infinite number of tunneling\nchannels under a finite gate voltage bias. A numerical check of the analytical\nsolution by means of Monte Carlo simulations has been performed recently. To\nderive the main result we study the so-called boundary parafermionic\nsine-Gordon model. The latter is of certain interest to condensed matter\napplications, namely as a toy model for a point junction in the multichannel\nquantum wire.\n", " Motivated by finding planar embeddings that lead to drawings with favorable\naesthetics, we study the problems MINMAXFACE and UNIFORMFACES of embedding a\ngiven biconnected multi-graph such that the largest face is as small as\npossible and such that all faces have the same size, respectively.\n We prove a complexity dichotomy for MINMAXFACE and show that deciding whether\nthe maximum is at most $k$ is polynomial-time solvable for $k \\leq 4$ and\nNP-complete for $k \\geq 5$. Further, we give a 6-approximation for minimizing\nthe maximum face in a planar embedding. For UNIFORMFACES, we show that the\nproblem is NP-complete for odd $k \\geq 7$ and even $k \\geq 10$. Moreover, we\ncharacterize the biconnected planar multi-graphs admitting 3- and 4-uniform\nembeddings (in a $k$-uniform embedding all faces have size $k$) and give an\nefficient algorithm for testing the existence of a 6-uniform embedding.\n", " Remarkable effectiveness of the channel or spatial attention mechanisms for\nproducing more discernible feature representation are illustrated in various\ncomputer vision tasks. However, modeling the cross-channel relationships with\nchannel dimensionality reduction may bring side effect in extracting deep\nvisual representations. In this paper, a novel efficient multi-scale attention\n(EMA) module is proposed. Focusing on retaining the information on per channel\nand decreasing the computational overhead, we reshape the partly channels into\nthe batch dimensions and group the channel dimensions into multiple\nsub-features which make the spatial semantic features well-distributed inside\neach feature group. Specifically, apart from encoding the global information to\nre-calibrate the channel-wise weight in each parallel branch, the output\nfeatures of the two parallel branches are further aggregated by a\ncross-dimension interaction for capturing pixel-level pairwise relationship. We\nconduct extensive ablation studies and experiments on image classification and\nobject detection tasks with popular benchmarks (e.g., CIFAR-100, ImageNet-1k,\nMS COCO and VisDrone2019) for evaluating its performance.\n", " The basic tool of classical results by Malkin and Melnikov on bifurcation of\nperiodic solutions from nondegenerate cycles of autonomous systems with\nperiodic perturbations is an implicit function theorem. In this paper the\nPoincare index is used to avoid the requirement of nondegeneracity for the\nunperturbed cycles and to provide additional geometrical properties of periodic\nsolutions of the perturbed system.\n", " Normal estimation on 3D point clouds is a fundamental problem in 3D vision\nand graphics. Current methods often show limited accuracy in predicting normals\nat sharp features (e.g., edges and corners) and less robustness to noise. In\nthis paper, we propose a novel normal estimation method for point clouds. It\nconsists of two phases: (a) feature encoding which learns representations of\nlocal patches, and (b) normal estimation that takes the learned representation\nas input and regresses the normal vector. We are motivated that local patches\non isotropic and anisotropic surfaces have similar or distinct normals, and\nthat separable features or representations can be learned to facilitate normal\nestimation. To realise this, we first construct triplets of local patches on 3D\npoint cloud data, and design a triplet network with a triplet loss for feature\nencoding. We then design a simple network with several MLPs and a loss function\nto regress the normal vector. Despite having a smaller network size compared to\nmost other methods, experimental results show that our method preserves sharp\nfeatures and achieves better normal estimation results on CAD-like shapes.\n", " We consider wave propagation and scattering governed by one dimensional\nSchrodinger operators with truncated periodic potentials. The propagation of\nwave packets with narrow frequency supports is studied. The goal is to describe\npotentials for which the group velocity (for periodic problem) is small and the\ntransmission coefficient for the truncated potential is not too small, i.e. to\nfind media where a slowing down of the wave packets coexists with a\ntransparency.\n", " Partially-observed Boolean dynamical systems (POBDS) are a general class of\nnonlinear models with application in estimation and control of Boolean\nprocesses based on noisy and incomplete measurements. The optimal minimum mean\nsquare error (MMSE) algorithms for POBDS state estimation, namely, the Boolean\nKalman filter (BKF) and Boolean Kalman smoother (BKS), are intractable in the\ncase of large systems, due to computational and memory requirements. To address\nthis, we propose approximate MMSE filtering and smoothing algorithms based on\nthe auxiliary particle filter (APF) method from sequential Monte-Carlo theory.\nThese algorithms are used jointly with maximum-likelihood (ML) methods for\nsimultaneous state and parameter estimation in POBDS models. In the presence of\ncontinuous parameters, ML estimation is performed using the\nexpectation-maximization (EM) algorithm; we develop for this purpose a special\nsmoother which reduces the computational complexity of the EM algorithm. The\nresulting particle-based adaptive filter is applied to a POBDS model of Boolean\ngene regulatory networks observed through noisy RNA-Seq time series data, and\nperformance is assessed through a series of numerical experiments using the\nwell-known cell cycle gene regulatory model.\n", " The Grassmann manifold G_{n,p}(L) is the set of all p-dimensional planes\n(through the origin) in the n-dimensional Euclidean space L^{n}, where L is\neither R or C. This paper considers an unequal dimensional quantization in\nwhich a source in G_{n,p}(L) is quantized through a code in G_{n,q}(L), where p\nand q are not necessarily the same. It is different from most works in\nliterature where p\\equiv q. The analysis for unequal dimensional quantization\nis based on the volume of a metric ball in G_{n,p}(L) whose center is in\nG_{n,q}(L). Our chief result is a closed-form formula for the volume of a\nmetric ball when the radius is sufficiently small. This volume formula holds\nfor Grassmann manifolds with arbitrary n, p, q and L, while previous results\npertained only to some special cases. Based on this volume formula, several\nbounds are derived for the rate distortion tradeoff assuming the quantization\nrate is sufficiently high. The lower and upper bounds on the distortion rate\nfunction are asymptotically identical, and so precisely quantify the asymptotic\nrate distortion tradeoff. We also show that random codes are asymptotically\noptimal in the sense that they achieve the minimum achievable distortion with\nprobability one as n and the code rate approach infinity linearly. Finally, we\ndiscuss some applications of the derived results to communication theory. A\ngeometric interpretation in the Grassmann manifold is developed for capacity\ncalculation of additive white Gaussian noise channel. Further, the derived\ndistortion rate function is beneficial to characterizing the effect of\nbeamforming matrix selection in multi-antenna communications.\n", " We consider the problem of assembling a sequence based on a collection of its\nsubstrings observed through a noisy channel. The mathematical basis of the\nproblem is the construction and design of sequences that may be discriminated\nbased on a collection of their substrings observed through a noisy channel. We\nexplain the connection between the sequence reconstruction problem and the\nproblem of DNA synthesis and sequencing, and introduce the notion of a DNA\nstorage channel. We analyze the number of sequence equivalence classes under\nthe channel mapping and propose new asymmetric coding techniques to combat the\neffects of synthesis and sequencing noise. In our analysis, we make use of\nrestricted de Bruijn graphs and Ehrhart theory for rational polytopes.\n", " In this work there are considered model problems for two nonlinear equations,\nwhich type depends on the solution. One of the equations may be called a\nnonlinear analog of the Lavrent'ev-Bitsadze equation.\n", " It has long been conjectured that generic dynamical systems has finite\nperiodic orbits, ever since the time of Poincar\\'e. In this article, a\nperturbation method is proposed for the $C^r$ closing of periodic orbits. This\nmethod is applicable to both time-varying and time-invariant vector fields.\n", " We study the periodic $q$-Whittaker and Hall-Littlewood processes, two\nprobability measures on sequences of partitions. We prove that a certain\nobservable of the periodic $q$-Whittaker process exhibits a $(q,u)$ symmetry\nafter a random shift, generalizing a previous result of Imamura, Mucciconi, and\nSasamoto who showed a matching between the periodic Schur and $q$-Whittaker\nmeasures, and also give a vertex model formulation of their result. As part of\nour proof of the $(q,u)$ symmetry, we obtain contour integral formulas for both\nthe periodic $q$-Whittaker and Hall-Littlewood processes. We also show a\nmatching between certain observables in the periodic Hall-Littlewood process\nand in a quasi-periodic stochastic six vertex model after a suitable random\nshift, and discuss a limit to the stationary periodic stochastic six vertex\nmodel.\n", " The rate for complete two-photon annihilation of molecular positronium\nPs$_{2}$ is reported. This decay channel involves a four-body collision among\nthe fermions forming Ps$_{2}$, and two photons of 1.022 MeV, each, as the final\nstate. The quantum electrodynamics result for the rate of this process is found\nto be $\\Gamma_{Ps_{2} \\rightarrow \\gamma\\gamma}$ = 9.0 $\\times 10^{-12}$\ns$^{-1}$. This decay channel completes the most comprehensive decay chart for\nPs$_{2}$ up to date.\n", " The embedding theorem arises in several problems from analysis and geometry.\nThe purpose of this paper is to provide a deeper understanding of analysis and\ngeometry with a particular focus on embedding theorems on spaces of homogeneous\ntype in the sense of Coifman and Weiss. We prove that embedding theorems hold\non spaces of homogeneous type if and only if geometric conditions, namely the\nmeasures of all balls have lower bounds, hold. As applications, our results\nprovide new and sharp previous related embedding theorems for the Sobolev,\nBesov and Triebel-Lizorkin spaces.\n", " For a general open set, we characterize the compactness of the embedding\n$W^{1,p}_0\\hookrightarrow L^q$ in terms of the summability of its torsion\nfunction. In particular, for $1\\le q<p$ we obtain that the embedding is\ncontinuous if and only if it is compact. The proofs crucially exploit a {\\it\ntorsional Hardy inequality} that we investigate in details.\n", " When the users in a MIMO broadcast channel experience different spatial\ntransmit correlation matrices, a class of gains is produced that is denoted\ntransmit correlation diversity. This idea was conceived for channels in which\ntransmit correlation matrices have mutually exclusive eigenspaces, allowing\nnon-interfering training and transmission. This paper broadens the scope of\ntransmit correlation diversity to the case of partially and fully overlapping\neigenspaces and introduces techniques to harvest these generalized gains. For\nthe two-user MIMO broadcast channel, we derive achievable degrees of freedom\n(DoF) and achievable rate regions with/without channel state information at the\nreceiver (CSIR). When CSIR is available, the proposed achievable DoF region is\ntight in some configurations of the number of receive antennas and the channel\ncorrelation ranks. We then extend the DoF results to the $K$-user case by\nanalyzing the interference graph that characterizes the overlapping structure\nof the eigenspaces. Our achievability results employ a combination of product\nsuperposition in the common part of the eigenspaces, and pre-beamforming (rate\nsplitting) to create multiple data streams in non-overlapping parts of the\neigenspaces. Massive MIMO is a natural example in which spatially correlated\nlink gains are likely to occur. We study the achievable downlink sum rate for a\nfrequency-division duplex massive MIMO system under transmit correlation\ndiversity.\n", " We establish a sharp Adams-type inequality invoking a Hardy inequality for\nany even dimension. This leads to a non compact Sobolev embedding in some\nOrlicz space. We also give a description of the lack of compactness of this\nembedding in the spirit of \\cite{Bahouri}.\n", " Nonuniformities in the imaging characteristics of modern image sensors are a\nprimary factor in the push to develop a pixel-level generalization of the\nphoton transfer characterization method. In this paper, we seek to develop a\nbody of theoretical results leading toward a comprehensive approach for\ntackling the biggest obstacle in the way of this goal: a means of pixel-level\nconversion gain estimation. This is accomplished by developing an estimator for\nthe reciprocal-difference of normal variances and then using this to construct\na novel estimator of the conversion gain. The first two moments of this\nestimator are derived and used to construct exact and approximate confidence\nintervals for its absolute relative bias and absolute coefficient of variation,\nrespectively. A means of approximating and computing optimal sample sizes are\nalso discussed and used to demonstrate the process of pixel-level conversion\ngain estimation for a real image sensor.\n", " It is observed that a natural analog of the Hahn-Banach theorem is valid for\nmetric functionals but fails for horofunctions. Several statements of the\nexistence of invariant metric functionals for individual isometries and\n1-Lipschitz maps are proved. Various other definitions, examples and facts are\npointed out related to this topic. In particular it is shown that the metric\n(horofunction) boundary of every infinite Cayley graphs contains at least two\npoints.\n", " In equilibrium molecular dynamics, Einstein relation can be used to calculate\nthe thermal conductivity. This method is equivalent to Green-Kubo relation and\nit does not require a derivation of an analytical form for the heat current.\nHowever, it is not commonly used as Green-Kubo relationship. Its wide use is\nhindered by the lack of a proper definition for integrated heat current (energy\nmoment) under periodic boundary conditions. In this paper, we developed an\nappropriate definition for integrated heat current to calculate thermal\nconductivity of solids under periodic conditions. We applied this method to\nsolid argon and silicon based systems; compared and contrasted with the\nGreen-Kubo approach.\n", " An L-shaped embedding of a tree in a point set is a planar drawing of the\ntree where the vertices are mapped to distinct points and every edge is drawn\nas a sequence of two axis-aligned line segments. There has been considerable\nwork on establishing upper bounds on the minimum cardinality of a point set to\nguarantee that any tree of the same size with maximum degree 4 admits an\nL-shaped embedding on the point set. However, no non-trivial lower bound is\nknown to this date, i.e., no known $n$-vertex tree requires more than $n$\npoints to be embedded. In this paper, we present the first examples of\n$n$-vertex trees for $n\\in\\{13,14,16,17,18,19,20\\}$ that require strictly more\npoints than vertices to admit an L-shaped embedding. Moreover, using computer\nhelp, we show that every tree on $n\\leq 12$ vertices admits an L-shaped\nembedding in every set of $n$ points. We also consider embedding ordered trees,\nwhere the cyclic order of the neighbors of each vertex in the embedding is\nprescribed. For this setting, we determine the smallest non-embeddable ordered\ntree on $n=10$ vertices, and we show that every ordered tree on $n\\leq 9$ or\n$n=11$ vertices admits an L-shaped embedding in every set of $n$ points. We\nalso construct an infinite family of ordered trees which do not always admit an\nL-shaped embedding, answering a question raised by Biedl, Chan, Derka, Jain,\nand Lubiw.\n", " A class of special solutions are constructed in an intuitive way for the\nultradiscrete analog of $q$-Painlev\\'e II ($q$-PII) equation. The solutions are\nclassified into four groups depending on the function-type and the system\nparameter.\n", " Movable antenna (MA) is a promising technology to improve wireless\ncommunication performance by varying the antenna position in a given finite\narea at the transceivers to create more favorable channel conditions. In this\npaper, we investigate the MA-enhanced multiple-access channel (MAC) for the\nuplink transmission from multiple users each equipped with a single MA to a\nbase station (BS) with a fixed-position antenna (FPA) array. A field-response\nbased channel model is used to characterize the multi-path channel between the\nantenna array of the BS and each user's MA with a flexible position. To\nevaluate the MAC performance gain provided by MAs, we formulate an optimization\nproblem for minimizing the total transmit power of users, subject to a\nminimum-achievable-rate requirement for each user, where the positions of MAs\nand the transmit powers of users, as well as the receive combining matrix of\nthe BS are jointly optimized. To solve this non-convex optimization problem\ninvolving intricately coupled variables, we develop two algorithms based on\nzero-forcing (ZF) and minimum mean square error (MMSE) combining methods,\nrespectively. Specifically, for each algorithm, the combining matrix of the BS\nand the total transmit power of users are expressed as a function of the MAs'\nposition vectors, which are then optimized by using the proposed\nmulti-directional descent (MDD) framework. It is shown that the proposed\nZF-based and MMSE-based MDD algorithms can converge to high-quality suboptimal\nsolutions with low computational complexities. Simulation results demonstrate\nthat the proposed solutions for MA-enhanced multiple access systems can\nsignificantly decrease the total transmit power of users as compared to\nconventional FPA systems employing antenna selection under both perfect and\nimperfect field-response information.\n", " We present a general method for the solution of the renormalization group\nequations for the non-forward parton distributions on the two-loop level in the\nflavour singlet channel based on an orthogonal polynomial reconstruction. Using\nthis formalism we study the effects of the evolution on recently proposed model\ndistribution functions.\n", " We present an illustrative application of the two famous mathematical\ntheorems in differential topology in order to show the existence of periodic\norbits with arbitrary given period for a class of hamiltonians .This result\npoint out for a mathematical answer for the long standing problem of existence\nof Planetary Sistems around stars.\n", " Orbit portraits were introduced by Milnor as a combinatorial tool to describe\nthe patterns of all periodic dynamical rays landing on a periodic cycle of a\nquadratic polynomial. This encodes information about the dynamics and the\nparameter spaces of these maps. We carry out a similar analysis for unicritical\nanti-polynomials and give an explicit description of the orbit portraits that\ncan occur for such maps in terms of their characteristic angles, which turns\nout to be rather restricted when compared with the holomorphic case. Finally,\nwe prove a realization theorem for these combinatorial objects. The results\nobtained in this paper serve as a combinatorial foundation for a detailed\nunderstanding of the combinatorics and topology of the parameter spaces of\nunicritical anti-polynomials and their connectedness loci, known as the\nmulticorns.\n", " In recent years, several algorithms for system identification with neural\nstate-space models have been introduced. Most of the proposed approaches are\naimed at reducing the computational complexity of the learning problem, by\nsplitting the optimization over short sub-sequences extracted from a longer\ntraining dataset. Different sequences are then processed simultaneously within\na minibatch, taking advantage of modern parallel hardware for deep learning. An\nissue arising in these methods is the need to assign an initial state for each\nof the sub-sequences, which is required to run simulations and thus to evaluate\nthe fitting loss. In this paper, we provide insights for calibration of neural\nstate-space training algorithms based on extensive experimentation and analyses\nperformed on two recognized system identification benchmarks. Particular focus\nis given to the choice and the role of the initial state estimation. We\ndemonstrate that advanced initial state estimation techniques are really\nrequired to achieve high performance on certain classes of dynamical systems,\nwhile for asymptotically stable ones basic procedures such as zero or random\ninitialization already yield competitive performance.\n", " Ethanediol is one of the largest complex organic molecules detected in space\nthus far. It has been found in different types of molecular clouds. The two\npropanediol isomers are the next larger diols. Hence, they are viable\ncandidates to be searched for in space. We wish to provide sufficiently large\nand accurate sets of spectroscopic parameters of 1,2-propanediol to facilitate\nsearches for this molecule at millimeter and longer submillimeter wavelengths.\nWe recorded rotational spectra of 1,2-propanediol in three wide frequency\nwindows between 38 and 400~GHz. We made extensive assignments for the three\nlowest energy conformers to yield spectroscopic parameters up to eighth order\nof angular momentum. Our present data will be helpful for identifying\n1,2-propanediol at moderate submillimeter or longer wavelengths with radio\ntelescope arrays such as ALMA, NOEMA, or EVLA. In particular, its detection\nwith ALMA in sources, in which ethanediol was detected, appears to be\npromising.\n", " For this year's 8TeV run of the LHC we lay out different strategies to search\nfor scalar top pairs. We show results for the hadronic and for the\nsemi-leptonic channels based on hadronic top tagging. For the di-lepton channel\nwe illustrate the impact of transverse mass variables. Each of our\nsignal-to-background ratios ranges around unity for a stop mass around 400GeV.\nThe combined signal significances show that dedicated stop searches are\nbecoming sensitive over a non-negligible part of parameter space.\n", " This paper reviews the main estimation and prediction results derived in the\ncontext of functional time series, when Hilbert and Banach spaces are\nconsidered, specially, in the context of autoregressive processes of order one\n(ARH(1) and ARB(1) processes, for H and B being a Hilbert and Banach space,\nrespectively). Particularly, we pay attention to the estimation and prediction\nresults, and statistical tests, derived in both parametric and non-parametric\nframeworks. A comparative study between different ARH(1) prediction approaches\nis developed in the simulation study undertaken.\n", " We study the impact of the ferroelectric domain wall on the ballistic\nconductance of the single-layer graphene channel in the heterostructure\ngraphene / physical gap / ferroelectric film using Wentzel-Kramers-Brillouin\napproximation. Both self-consistent numerical modeling of the electric field\nand space charge dynamics in the heterostructure and approximate analytical\ntheory show that the domain wall contact with the surface creates p-n junction\nin graphene channel. We calculated that the carriers' concentration induced in\ngraphene by uncompensated ferroelectric dipoles originated from the spontaneous\npolarization abrupt near the surface can reach the values of 1019 m-2 order,\nwhich is in two orders higher than it can be obtained for the gate doped\ngraphene on non-ferroelectric substrates. Therefore we predict that graphene\nchannel with the p-n junction caused by ferroelectric domain wall would be\ncharacterized by rather high ballistic conductivity.\n", " This paper deals with the stability of linear periodic difference delay\nsystems, where the value at time $t$ of a solution is a linear combination with\nperiodic coefficients of its values at finitely many delayed instants\n$t-\\tau_1,\\ldots,t-\\tau_N$. We establish a necessary and sufficient condition\nfor exponential stability of such systems when the coefficients have\nH\\\"older-continuous derivative, that generalizes the one obtained for\ndifference delay systems with constant coefficients by Henry and Hale in the\n1970s. This condition may be construed as analyticity, in a half plane, of the\n(operator valued) harmonic transfer function of an associated linear control\nsystem.\n", " We present a $(1- \\varepsilon)$-approximation algorithms for maximum\ncardinality matchings in disk intersection graphs -- all with near linear\nrunning time. We also present estimation algorithm that returns $(1\\pm\n\\varepsilon)$-approximation to the size of such matchings -- this algorithms\nrun in linear time for unit disks, and $O(n \\log n)$ for general disks (as long\nas the density is relatively small).\n", " We point out an error in the proof of the main result of the paper of Tanabe\net al. (2007) concerning a parameter estimation for von Mises--Fisher\ndistributions, we correct the proof of the main result and we present a short\nalternative proof.\n", " Results obtained with multichannel installation created for long-term studies\nof various processes, are collated with the data published by J.H. Jenkins and\nE.Fischbach, who found a decrease of 54Mn radioactivity near the time of series\nof solar flares between 5 and 17 December 2006. Analysis of the data from our\ninstallation in December 2006 has not revealed any deviations from the usual\nbehaviour of the count rates for 90Sr-90Y, 60Co and 239Pu sources. The same can\nbe said of the data collected during the period of highly powerful solar flares\nbetween 19 October and 4 November 2003. Apparent drops in the count rate were\ndetected between 10 and 12 May 2002 while registering the activity of 60Co and\non 19 and 20 June 2004 for 90Sr-90Y source. Around the time of these events, no\nobservations of large solar flares were reported. Thus, proposed link between\nthe drop in the rates of radioactive decay and appearance of solar flares could\nnot be confirmed. From obtained outcomes follows, that the radioactivity drop\neffect, if it really exists, is rather rare, and that the reason calling this\neffect unequally influences various radioactive sources.\n" ]

[ " Counterfactual explanations (CFEs) highlight what changes to a model's input\nwould have changed its prediction in a particular way. CFEs have gained\nconsiderable traction as a psychologically grounded solution for explainable\nartificial intelligence (XAI). Recent innovations introduce the notion of\ncomputational plausibility for automatically generated CFEs, enhancing their\nrobustness by exclusively creating plausible explanations. However, practical\nbenefits of such a constraint on user experience and behavior is yet unclear.\nIn this study, we evaluate objective and subjective usability of\ncomputationally plausible CFEs in an iterative learning design targeting novice\nusers. We rely on a novel, game-like experimental design, revolving around an\nabstract scenario. Our results show that novice users actually benefit less\nfrom receiving computationally plausible rather than closest CFEs that produce\nminimal changes leading to the desired outcome. Responses in a post-game survey\nreveal no differences in terms of subjective user experience between both\ngroups. Following the view of psychological plausibility as comparative\nsimilarity, this may be explained by the fact that users in the closest\ncondition experience their CFEs as more psychologically plausible than the\ncomputationally plausible counterpart. In sum, our work highlights a\nlittle-considered divergence of definitions of computational plausibility and\npsychological plausibility, critically confirming the need to incorporate human\nbehavior, preferences and mental models already at the design stages of XAI\napproaches. In the interest of reproducible research, all source code, acquired\nuser data, and evaluation scripts of the current study are available:\nhttps://github.com/ukuhl/PlausibleAlienZoo\n", " The formation of trapped surfaces in the head-on collision of shock waves in\nconformal and non-conformal backgrounds is investigated. The backgrounds in-\nclude all interesting confining and non-confining backgrounds that may be\nrelevant for QCD. Several transverse profiles of the shocks are investigated\nincluding distributions that fall-off as powers or exponentials. Different ways\nof cutting-off the UV contributions (that are expected to be perturbative in\nQCD) are explored. Under some plausible simplifying assumptions our estimates\nare converted into predictions for multiplicities for heavy-ion collisions at\nRHIC and LHC.\n", " Triangular mass matrices for neutrinos and their charged partners contain\nfull information on neutrino mixing in a most concise form. Although the scheme\nis general and model independent, triangular matrices are typical for reducible\nbut indecomposable representations of graded Lie algebras which, in turn, are\ncharacteristic for the standard model in noncommutative geometry. The mixing\nmatrix responsible for neutrino oscillations is worked out analytically for two\nand three lepton families. The example of two families fixes the mixing angle\nto just about what is required by the Mikheyev-Smirnov-Wolfenstein resonance\noscillation of solar neutrinos. In the case of three families we classify all\nphysically plausible choices for the neutrino mass matrix and derive\ninteresting bounds on some of the moduli of the mixing matrix.\n", " We consider a class of supersymmetric models containing baryon number\nviolating processes such as observable neutron - antineutron oscillations that\nare mediated by color triplet diquark fields. For plausible values of the\ndiquark-quark couplings, the scalar diquark with mass between a few hundred GeV\nand one TeV or so can be produced in the s-channel at the LHC and detected\nthrough its decay into a top quark and a hadronic jet.\n", " Some solutions to the anomalies of ultra high energy cosmic-ray(UHECR) and\nTeV $\\gamma$-rays require disturbed non-quadratic dispersion relations, which\nsuggest the Lorentz violation. Also, some new theories such as quantum gravity,\nstring theory and the standard model extension imply the Lorentz violation. In\nthis paper, we derive some transformation laws for the classical parameters of\nnonlinear field system, and then provide their dispersion relations. These\ndispersion relations also have non-quadratic form but keep the Lorentz\ninvariance. The analysis of this paper may be helpful for understanding the\nquantum theory and the plausible Lorentz violation.\n", " A new kinematic condition for soliton motions of an $n$-dimensional continuum\nin $R^{n+m}$, independent of the underlying physics, is proven. The condition\nand its consequences for different cases are demonstrated. A soliton in a 1D\nstring that rocks back and forth, a rotating soliton in a 2D membrane, and\nvarious other cases are presented as examples. It is shown that traveling knots\nbased on classical wave equation are plausible. Cases in which all the motions\nare solitons are also presented. Compatibility of equations of motion with the\nkinematic constraint is explored and demonstrated.\n", " Recently, a quantum mechanical theory of quantum spaces described by a large\n$N$ non-commutative coordinates is proposed as a theory for quantum gravity\n[1]. In this paper, we construct Kerr black hole as a rotating noncommutative\ngeometry solution of this theory. Due to rotation, the fuzzy sphere is deformed\ninto a fuzzy ellipsoid, which matches exactly the outer horizon of the Kerr\nblack hole in the Boyer-Lindquist coordinates. Together with a half-filled\nFermi sea, the fuzzy solution reproduces the Bekenstein-Hawking entropy as well\nas the mass of the Kerr black hole. These results provide further support that\nthe proposed theory of quantum spaces is a plausible candidate for the theory\nof quantum gravity.\n", " Solar activity It is vital to accurately predict solar activity, in order to\ndecrease the plausible damage of electronic equipment in the event of a large\nhigh-intensity solar eruption. Recently, we have proposed BELFIS (Brain\nEmotional Learning-based Fuzzy Inference System) as a tool for the forecasting\nof chaotic systems. The structure of BELFIS is designed based on the neural\nstructure of fear conditioning. The function of BELFIS is implemented by\nassigning adaptive networks to the components of the BELFIS structure. This\npaper especially focuses on performance evaluation of BELFIS as a predictor by\nforecasting solar cycles 16 to 24. The performance of BELFIS is compared with\nother computational models used for this purpose, and in particular with\nadaptive neuro-fuzzy inference system (ANFIS).\n", " The amount of published research in Physics Education Research (PER) shows,\non one hand, an increasing interest in the design and development of high\nperformance physics teaching strategies, and, on the other hand, it tries to\nunderstand plausible ways on which the brain processes scientific information\nso that scientific thinking skills could be taught more effectively. As physics\nis a subject in which mathematical and conceptual reasoning can not be\nseparated, instructors of physics face the problem of finding suitable advise\non the most effective methods of teaching physics (i.e. how much time should be\nspent on intuitive conceptual reasoning and how much time in developing\nquantitative reasoning, and how to teach both in a unitary way). In spite the\nimportant efforts made by the PER community, the published results are\noverwhelming and confusing for the physics instructors in the sense that the\nconclusions that have arisen in those articles are in some instances\ncontroversial and far from being conclusive in pointing out a particular\nstrategy to overcome the afore mentioned problem. Accordingly, based on the\nanalysis of published PER work, we'll be arguing that one of the major\ndifficulties to overcome in the teaching of physics could be associated to the\nlack of a consistent and coherent methodological framework for teaching which\nintegrates both aspects, conceptual and mathematical reasoning, in a systemic\nway of thinking. We will be presenting a set of plausible steps that could be\napplied to tackle the aforementioned difficulty.\n", " Systems biology and whole-cell modelling are demanding increasingly\ncomprehensive mathematical models of cellular biochemistry. These models\nrequire the development of simplified models of specific processes which\ncapture essential biophysical features but without unnecessarily complexity.\nRecently there has been renewed interest in thermodynamically-based modelling\nof cellular processes. Here we present an approach to developing of simplified\nyet thermodynamically consistent (hence physically plausible) models which can\nreadily be incorporated into large scale biochemical descriptions but which do\nnot require full mechanistic detail of the underlying processes. We illustrate\nthe approach through development of a simplified, physically plausible model of\nthe mitochondrial electron transport chain and show that the simplified model\nbehaves like the full system.\n", " In the wake of the burgeoning expansion of generative artificial intelligence\n(AI) services, the computational demands inherent to these technologies\nfrequently necessitate cloud-powered computational offloading, particularly for\nresource-constrained mobile devices. These services commonly employ prompts to\nsteer the generative process, and both the prompts and the resultant content,\nsuch as text and images, may harbor privacy-sensitive or confidential\ninformation, thereby elevating security and privacy risks. To mitigate these\nconcerns, we introduce $\\Lambda$-Split, a split computing framework to\nfacilitate computational offloading while simultaneously fortifying data\nprivacy against risks such as eavesdropping and unauthorized access. In\n$\\Lambda$-Split, a generative model, usually a deep neural network (DNN), is\npartitioned into three sub-models and distributed across the user's local\ndevice and a cloud server: the input-side and output-side sub-models are\nallocated to the local, while the intermediate, computationally-intensive\nsub-model resides on the cloud server. This architecture ensures that only the\nhidden layer outputs are transmitted, thereby preventing the external\ntransmission of privacy-sensitive raw input and output data. Given the\nblack-box nature of DNNs, estimating the original input or output from\nintercepted hidden layer outputs poses a significant challenge for malicious\neavesdroppers. Moreover, $\\Lambda$-Split is orthogonal to traditional\nencryption-based security mechanisms, offering enhanced security when deployed\nin conjunction. We empirically validate the efficacy of the $\\Lambda$-Split\nframework using Llama 2 and Stable Diffusion XL, representative large language\nand diffusion models developed by Meta and Stability AI, respectively. Our\n$\\Lambda$-Split implementation is publicly accessible at\nhttps://github.com/nishio-laboratory/lambda_split.\n", " The framework of cognitively bounded rationality treats problem solving as\nfundamentally rational, but emphasises that it is constrained by cognitive\narchitecture and the task environment. This paper investigates a simple\ndecision making heuristic, Take The Best (TTB), within that framework. We\nformulate TTB as a likelihood-based probabilistic model, where the decision\nstrategy arises by probabilistic inference based on the training data and the\nmodel constraints. The strengths of the probabilistic formulation, in addition\nto providing a bounded rational account of the learning of the heuristic,\ninclude natural extensibility with additional cognitively plausible constraints\nand prior information, and the possibility to embed the heuristic as a subpart\nof a larger probabilistic model. We extend the model to learn cue\ndiscrimination thresholds for continuous-valued cues and experiment with using\nthe model to account for biased preference feedback from a boundedly rational\nagent in a simulated interactive machine learning task.\n", " The hydrodynamic description of the Fermi arc surface states is proposed. In\nview of the strong suppression of scattering on impurities, the hydrodynamic\nregime for Fermi arc states should be, in principle, plausible. By using the\nkinetic theory, the Fermi arc hydrodynamics is derived and the corresponding\neffects on the bulk flow and surface collective modes are studied. For the bulk\nflow, the key effect of the proposed Fermi arc hydrodynamics is the\nmodification of the corresponding boundary conditions. In a slab geometry, it\nis shown that, depending on the transfer rates between the surface and bulk,\nthe hydrodynamic flow of the electron fluid inside the slab could be\nsignificantly altered and even enhanced near the surfaces. As to the spectrum\nof the surface collective modes, in agreement with earlier studies, it is found\nthat the Fermi arcs allow for an additional gapless spectrum branch and a\nstrong anisotropy of the surface plasmon dispersion relations in momentum\nspace. The gapped modes are characterized by closed elliptic contours of\nconstant frequency in momentum space.\n", " Recent studies have reported that T cells can integrate signals between\ninterrupted encounters with Antigen Presenting Cells (APCs) in such a way that\nthe process of signal integration exhibits a form of memory. Here, we carry out\na computational study using a simple mathematical model of T cell activation to\ninvestigate the ramifications of interrupted T cell-APC contacts on signal\nintegration. We consider several mechanisms of how signal integration at these\ntime scales may be achieved and conclude that feedback control of immediate\nearly gene products (IEGs) appears to be a highly plausible mechanism that\nallows for effective signal integration and cytokine production from multiple\nexposures to APCs. Analysis of these computer simulations provides an\nexperimental roadmap involving several testable predictions.\n", " The properties of ground-state baryons of light-quarks are investigated in\nU(12)-classification scheme of hadrons, recently proposed by us. In U(12), in\naddition to the ordinary 56 of SU(6)SF, the existence of the extra 56 with\npositive-parity and 70 with negative-parity appear as ground-states in lower\nmass region. The N(1440), Lambda(1600) and Sigma(1660) have the plausible\nproperties of masses and strong decay widths as the flavor-octet of extra 56,\nwhile the ordinary radially-excited octet states are expected not to be\nobserved as resonances because of their large predicted widths of the decays to\nextra 70 baryons. The extra 70 baryons are not observed directly as resonances\nfor the same reason, except for only the Lambda(1406), of which properties are\nconsistently explained through the singlet-octet mixing. The baryon properties\nin lower mass regions are consistently explained in U(12) scheme.\n", " Here, we propose a novel approach to experimentally and theoretically study\nthe properties of QCD matter under new extreme conditions, namely having an\ninitial temperature over 300~MeV and baryonic charge density over three times\nthe values of the normal nuclear density. According to contemporary theoretical\nknowledge, such conditions were not accessible during the early Universe\nevolution and are not accessible now in the known astrophysical phenomena. To\nachieve these new extreme conditions, we proposed performing high-luminosity\nexperiments at LHC or other colliders by means of scattering the two colliding\nbeams at the nuclei of a solid target that is fixed at their interaction\nregion. Under plausible assumptions, we estimate the reaction rate for the\np+C+p and Pb+Pb+Pb reactions and discuss the energy deposition into the target\nand possible types of fixed targets for such reactions. To simulate the triple\nnuclear collisions, we employed the well-known UrQMD 3.4 model for the beam\ncenter-of-mass collision energies 2.76 TeV. As a result of our modeling, we\nfound that, in the most central and simultaneous triple nuclear collisions, the\ninitial baryonic charge density is approximately three times higher than the\none achieved in the ordinary binary nuclear collisions at this energy.\n", " Understanding the nature of time remains a key unsolved problem in science.\nNewton in the Principia asserted an absolute universal time that {\\it `flows\nequably'}. Hamilton then proposed a mathematical unification of space and time\nwithin the framework of the quaternions that ultimately lead to the famous\nMinkowski formulation in 1908 using four-vectors. The Minkowski framework is\nfound to provide a versatile formalism for describing the relationship between\nspace and time in accordance with relativistic principles, but nevertheless\nfails to provide deeper insights into the physical origin of time and its\nproperties. In this paper we begin with a recognition of the fundamental role\nplayed by three-dimensional space in physics that we model using the Clifford\nalgebra multivector. From this geometrical foundation we are then able to\nidentify a plausible origin for our concept of time. This geometrical\nperspective also allows us to make a key topological distinction between time\nand space, with time being a point-like quantity. The multivector then allows a\ngeneralized unification of time and space within a Minkowski-like description.\n", " String theory posesses numerous axion candidates. The recent realization that\nthe compactification radius in string theory might be large means that these\nstates can solve the strong CP problem. This still leaves the question of the\ncosmological bound on the axion mass. Here we explore two schemes for\naccommodating such light axions in cosmology. In the first, we note that in\nstring theory the universe is likely to be dominated early on by the coherent\noscillations of some moduli. The usual moduli problem assumes that these fields\nhave masses comparable to the gravitino. We argue that string moduli are likely\nto be substantially more massive, eliminating this problem. In such cosmologies\nthe axion bound is significantly weakened. Plausible mechanisms for generating\nthe baryon number density are described. In the second, we point out that in\nstring theory, the axion potentials might be much larger at early times than at\npresent. In string theory, if CP violation is described by a small parameter,\nthe axion may sit sufficiently close to its true minimum to invalidate the\nbounds.\n", " Studying the function spaces defined by neural networks helps to understand\nthe corresponding learning models and their inductive bias. While in some\nlimits neural networks correspond to function spaces that are reproducing\nkernel Hilbert spaces, these regimes do not capture the properties of the\nnetworks used in practice. In contrast, in this paper we show that deep neural\nnetworks define suitable reproducing kernel Banach spaces.\n These spaces are equipped with norms that enforce a form of sparsity,\nenabling them to adapt to potential latent structures within the input data and\ntheir representations. In particular, leveraging the theory of reproducing\nkernel Banach spaces, combined with variational results, we derive representer\ntheorems that justify the finite architectures commonly employed in\napplications. Our study extends analogous results for shallow networks and can\nbe seen as a step towards considering more practically plausible neural\narchitectures.\n", " This paper addresses the problem of generating 3D interactive human motion\nfrom text. Given a textual description depicting the actions of different body\nparts in contact with objects, we synthesize sequences of 3D body poses that\nare visually natural and physically plausible. Yet, this task poses a\nsignificant challenge due to the inadequate consideration of interactions by\nphysical contacts in both motion and textual descriptions, leading to unnatural\nand implausible sequences. To tackle this challenge, we create a novel dataset\nnamed RICH-CAT, representing ``Contact-Aware Texts'' constructed from the RICH\ndataset. RICH-CAT comprises high-quality motion, accurate human-object contact\nlabels, and detailed textual descriptions, encompassing over 8,500 motion-text\npairs across 26 indoor/outdoor actions. Leveraging RICH-CAT, we propose a novel\napproach named CATMO for text-driven interactive human motion synthesis that\nexplicitly integrates human body contacts as evidence. We employ two VQ-VAE\nmodels to encode motion and body contact sequences into distinct yet\ncomplementary latent spaces and an intertwined GPT for generating human motions\nand contacts in a mutually conditioned manner. Additionally, we introduce a\npre-trained text encoder to learn textual embeddings that better discriminate\namong various contact types, allowing for more precise control over synthesized\nmotions and contacts. Our experiments demonstrate the superior performance of\nour approach compared to existing text-to-motion methods, producing stable,\ncontact-aware motion sequences. Code and data will be available for research\npurposes.\n", " We investigate the possible sequels, for the hybrid compact stars consisting\nof nucleons, hyperons and three flavour color-flavor-locked quarks phase under\nglobal neutrality and chemical equilibrium conditions. The hadron equations of\nstate are computed within the framework of energy density functionals based on\nthe relativistic mean field theory by employing two different model\nparameterizations. The quarks matter phase of equation of state is computed by\nusing Quarks Quasiparticle model derived from a non-relativistic energy\ndensity-functional approach. A plausible set of hybrid equations of state for\nsuperdense hadron-quarks matter is obtained and, which satisfies the\nconstraints provided by bulk nuclear matter, the observational data of\nfrequencies and maximum Gravitational mass 2$M_\\odot$ with their extracted\nradii, Keplerian limits, secular axisymmetric instability and limits of tidal\ndeformabilities from GW170817 binary neutron stars (BNS) merger. We obtained\nstructural properties of nonrotating and rotating compact stars, the\nevolutionary sequences with constant baryonic mass star spinning down by\nelectromagnetic and gravitational radiations and, discuss the issues relating\nto phase transition from hadrons to quarks matter phase. The internal structure\nof rotating star with observed spin down frequencies, exhibiting shrinkage of\nsoft quarks core of compact stars are discussed for constant baryonic mass. We\npresent the theoretically computed limits of radii for the spin down\nconfigurations of hybrid stars corresponding to the recently observed\nmillisecond pulsars. We also present and discuss the calculated results for\ntidal deformability, $\\Lambda$ and tidal parameters $\\Lambda_1$, $\\Lambda_2$ of\nnon spining hybrid star which are consistent with the waveform model analysis\nof GW170817 BNS.\n", " In a recent Foundations of Physics paper [5] by the current author it was\nshown that, when the self-force problem of classical electrodynamics is\nproperly solved, it becomes a plausible ontology underlying the statistical\ndescription of quantum mechanics. In the current paper we extend this result,\nshowing that ordinary matter, thus represented, possibly suffices in explaining\nthe outstanding observations currently requiring for this task the contrived\nnotions of dark-matter, dark-energy and inflation. The single mandatory `fix'\nto classical electrodynamics, demystifying both very small and very large scale\nphysics, should be contrasted with other ad hoc solutions to either problems.\nInstrumental to our cosmological model is scale covariance (and `spontaneous\nbreaking' thereof), a formal symmetry of classical electrodynamics treated on\nequal footing with its Poincare covariance, which is incompatible with the\n(absolute) metrical attributes of the GR metric tensor.\n", " We revisit the classic task of finding the shortest tour of $n$ points in\n$d$-dimensional Euclidean space, for any fixed constant $d \\geq 2$. We\ndetermine the optimal dependence on $\\varepsilon$ in the running time of an\nalgorithm that computes a $(1+\\varepsilon)$-approximate tour, under a plausible\nassumption. Specifically, we give an algorithm that runs in\n$2^{O(1/\\varepsilon^{d-1})} n\\log n$ time. This improves the previously\nsmallest dependence on $\\varepsilon$ in the running time\n$(1/\\varepsilon)^{O(1/\\varepsilon^{d-1})}n \\log n$ of the algorithm by Rao and\nSmith (STOC 1998). We also show that a\n$2^{o(1/\\varepsilon^{d-1})}\\mathrm{poly}(n)$ algorithm would violate the\nGap-Exponential Time Hypothesis (Gap-ETH).\n Our new algorithm builds upon the celebrated quadtree-based methods initially\nproposed by Arora (J. ACM 1998), but it adds a new idea that we call\nsparsity-sensitive patching. On a high level this lets the granularity with\nwhich we simplify the tour depend on how sparse it is locally. We demonstrate\nthat our technique extends to other problems, by showing that for Steiner Tree\nand Rectilinear Steiner Tree it yields the same running time. We complement our\nresults with a matching Gap-ETH lower bound for Rectilinear Steiner Tree.\n", " Temporal Graph Learning, which aims to model the time-evolving nature of\ngraphs, has gained increasing attention and achieved remarkable performance\nrecently. However, in reality, graph structures are often incomplete and noisy,\nwhich hinders temporal graph networks (TGNs) from learning informative\nrepresentations. Graph contrastive learning uses data augmentation to generate\nplausible variations of existing data and learn robust representations.\nHowever, rule-based augmentation approaches may be suboptimal as they lack\nlearnability and fail to leverage rich information from downstream tasks. To\naddress these issues, we propose a Time-aware Graph Structure Learning (TGSL)\napproach via sequence prediction on temporal graphs, which learns better graph\nstructures for downstream tasks through adding potential temporal edges. In\nparticular, it predicts time-aware context embedding based on previously\nobserved interactions and uses the Gumble-Top-K to select the closest candidate\nedges to this context embedding. Additionally, several candidate sampling\nstrategies are proposed to ensure both efficiency and diversity. Furthermore,\nwe jointly learn the graph structure and TGNs in an end-to-end manner and\nperform inference on the refined graph. Extensive experiments on temporal link\nprediction benchmarks demonstrate that TGSL yields significant gains for the\npopular TGNs such as TGAT and GraphMixer, and it outperforms other contrastive\nlearning methods on temporal graphs. We release the code at\nhttps://github.com/ViktorAxelsen/TGSL.\n", " We argue that both the positron fraction measured by PAMELA and the peculiar\nspectral features reported in the total electron-positron (e+e-) flux measured\nby ATIC have a very natural explanation in electron-positron pairs produced by\nnearby pulsars. While this possibility was pointed out a long time ago, the\ngreatly improved quality of current data potentially allow to reverse-engineer\nthe problem: given the regions of pulsar parameter space favored by PAMELA and\nby ATIC, are there known pulsars that explain the data with reasonable\nassumptions on the injected e+e- pairs? In the context of simple benchmark\nmodels for estimating the e+e- output, we consider all known pulsars, as listed\nin the most complete available catalogue. We find that it is unlikely that a\nsingle pulsar be responsible for both the PAMELA e+ fraction anomaly and for\nthe ATIC excess, although two single sources are in principle enough to explain\nboth experimental results. The PAMELA excess e+ likely come from a set of\nmature pulsars (age ~ 10^6 yr), with a distance of 0.8-1 kpc, or from a single,\nyounger and closer source like Geminga. The ATIC data require a larger (and\nless plausible) energy output, and favor an origin associated to powerful, more\ndistant (1-2 kpc) and younger (age ~ 10^5$ yr) pulsars. We list several\ncandidate pulsars that can individually or coherently contribute to explain the\nPAMELA and ATIC data. Although generally suppressed, we find that the\ncontribution of pulsars more distant than 1-2 kpc could contribute for the ATIC\nexcess. Finally, we stress the multi-faceted and decisive role that Fermi-LAT\nwill play in the very near future by (1) providing an exquisite measurement of\nthe e+e- flux, (2) unveiling the existence of as yet undetected pulsars, and\n(3) searching for anisotropies in the arrival direction of high-energy e+e-.\n", " Goal-oriented dialog has been given attention due to its numerous\napplications in artificial intelligence. Goal-oriented dialogue tasks occur\nwhen a questioner asks an action-oriented question and an answerer responds\nwith the intent of letting the questioner know a correct action to take. To ask\nthe adequate question, deep learning and reinforcement learning have been\nrecently applied. However, these approaches struggle to find a competent\nrecurrent neural questioner, owing to the complexity of learning a series of\nsentences. Motivated by theory of mind, we propose \"Answerer in Questioner's\nMind\" (AQM), a novel information theoretic algorithm for goal-oriented dialog.\nWith AQM, a questioner asks and infers based on an approximated probabilistic\nmodel of the answerer. The questioner figures out the answerer's intention via\nselecting a plausible question by explicitly calculating the information gain\nof the candidate intentions and possible answers to each question. We test our\nframework on two goal-oriented visual dialog tasks: \"MNIST Counting Dialog\" and\n\"GuessWhat?!\". In our experiments, AQM outperforms comparative algorithms by a\nlarge margin.\n", " Strange Quark Nuggets are the relics of the microsecond old universe after\nthe big bang. The universe having experienced a mini inflation of 7-e folding\nhas gone through supercooling leading to a first order phase transition from\nquark to hadrons. Strange Quark Nuggets, the relics of this phase transition\nwill constitute 12.5% or at the most 25% of the dark matter of the universe.\nOnly 10-3 or 10-4 part of the total number of SQNs form binaries, and only some\nof the binaries will turn to black holes. Primordial Black Hole (PBH) as a\nrelic is shown to be not a plausible scenario. Simultaneous observation of high\nenergy gamma rays from SQN binaries formed from coalescence of SQNs along with\nblack holes will be the vital cursor of this evolutionary scenario just\npresented.\n", " The unification of Quantum Mechanics and General Relativity remains the\nprimary goal of Theoretical Physics, with string theory appearing as the only\nplausible unifying scheme. In the present work, in a search of the conceptual\nfoundations of string theory, we analyze the relational logic developed by C.\nS. Peirce in the late nineteenth century. The Peircean logic has the\nmathematical structure of a category with the relation $R_{ij}$ among two\nindividual terms $S_i$ and $S_j$, serving as an arrow (or morphism). We\nintroduce a realization of the corresponding categorical algebra of\ncompositions, which naturally gives rise to the fundamental quantum laws, thus\nunderscoring the relational character of Quantum Mechanic. The same relational\nalgebra generates a number of group structures, among them $W_{\\infty}$. The\ngroup $W_{\\infty}$ is embodied and realized by the matrix models, themselves\nclosely linked with string theory. It is suggested that relational logic and in\ngeneral category theory may provide a new paradigm, within which to develop\nmodern physical theories.\n", " We produce combinatorial models for configuration space in a simplicial\ncomplex, and for configurations near a single point (\"local configuration\nspace.\") The model for local configuration space is built out of the poset of\nposet structures on a finite set. The model for global configuration space\nrelies on a combinatorial model for a simplicial complex with a deleted\nsubcomplex. By way of application, we study the nodal curve $y^2 z = x^3 + x^2\nz$, obtaining a presentation for its two-strand braid group, a conjectural\npresentation for its three-strand braid group, and presentations for its two-\nand three-strand local braid groups near the singular point.\n", " We prove an almost everywhere convergence result for Bochner-Riesz means of\n$L^p$ functions on Heisenberg-type groups, yielding the existence of a $p>2$\nfor which convergence holds for means of arbitrarily small order. The proof\nhinges on a reduction of weighted $L^2$ estimates for the maximal Bochner-Riesz\noperator to corresponding estimates for the non-maximal operator, and a `dual\nSobolev trace lemma', whose proof is based on refined estimates for Jacobi\npolynomials.\n", " The paper discusses sharp sufficient conditions for interpolation and\nsampling for functions of n variables with convex spectrum. When n=1, the\nclassical theorems of Ingham and Beurling state that the critical values in the\nestimates from above (from below) for the distances between interpolation\n(sampling) nodes are the same. This is no longer true for n>1. While the\ncritical value for sampling sets remains constant, the one for interpolation\ngrows linearly with the dimension.\n", " For this year's 8TeV run of the LHC we lay out different strategies to search\nfor scalar top pairs. We show results for the hadronic and for the\nsemi-leptonic channels based on hadronic top tagging. For the di-lepton channel\nwe illustrate the impact of transverse mass variables. Each of our\nsignal-to-background ratios ranges around unity for a stop mass around 400GeV.\nThe combined signal significances show that dedicated stop searches are\nbecoming sensitive over a non-negligible part of parameter space.\n", " A generalised trapezoid inequality for convex functions and applications for\nquadrature rules are given. A refinement and a counterpart result for the\nHermite-Hadamard inequalities are obtained and some inequalities for pdf's and\n(HH)-divergence measure are also mentioned.\n", " Given a finite family of squares in the plane, the packing problem asks for\nthe maximum number $\\nu$ of pairwise disjoint squares among them, while the\nhitting problem for the minimum number $\\tau$ of points hitting all of them.\nClearly, $\\tau \\ge \\nu$. Both problems are known to be NP-hard, even for\nfamilies of axis-parallel unit squares.\n The main results of this work provide the first non-trivial bounds for the\n$\\tau / \\nu$ ratio for not necessarily axis-parallel squares. We establish an\nupper bound of $6$ for unit squares and $10$ for squares of varying sizes. The\nworst ratios we can provide with examples are $3$ and $4$, respectively. For\ncomparison, in the axis-parallel case, the supremum of the considered ratio is\nin the interval $[\\frac{3}{2},2]$ for unit squares and $[\\frac{3}{2},4]$ for\nsquares of varying sizes. The methods we introduced for the $\\tau/\\nu$ ratio\ncan also be used to relate the chromatic number $\\chi$ and clique number\n$\\omega$ of squares by bounding the $\\chi/\\omega$ ratio by $6$ for unit squares\nand $9$ for squares of varying sizes.\n The $\\tau / \\nu$ and $\\chi/\\omega$ ratios have already been bounded before by\na constant for \"fat\" objects, the fattest and simplest of which are disks and\nsquares. However, while disks have received significant attention, specific\nbounds for squares have remained essentially unexplored. This work intends to\nfill this gap.\n", " Coulomb's constant is defined as a value for attraction and repulsion.\nHowever, it is strange that only one value can be applied for both attraction\nand repulsion. A very little difference between coulomb's constant for\nattraction and coulomb's constant for repulsion can be the source of gravity.\nThe author verified if that theory is correct by calculating with slightly\nbigger coulomb's constant for attraction.\n", " In our recent paper we characterized the extension property for symmetrized\nweight composition for linear codes over a module alphabet. Several\nimprovements for the case of vector space alphabets are given in this paper. A\ndetailed description of the property of $G$-pseudo-injectivity for vector\nspaces is made.\n", " Understanding the skill sets and knowledge required for any career is of\nutmost importance, but it is increasingly challenging in today's dynamic world\nwith rapid changes in terms of the tools and techniques used. Thus, it is\nespecially important to be able to accurately identify the required skill sets\nfor any job for better career insights and development. In this paper, we\npropose and develop the Skill Recommendation (SkillRec) system for recommending\nthe relevant job skills required for a given job based on the job title.\nSkillRec collects and identify the skill set required for a job based on the\njob descriptions published by companies hiring for these roles. In addition to\nthe data collection and pre-processing capabilities, SkillRec also utilises\nword/sentence embedding techniques for job title representation, alongside a\nfeed-forward neural network for job skill recommendation based on the job title\nrepresentation. Based on our preliminary experiments on a dataset of 6,000 job\ntitles and descriptions, SkillRec shows a promising performance in terms of\naccuracy and F1-score.\n", " A Neumann series of Bessel functions (NSBF) representation for solutions of\nSturm-Liouville equations and for their derivatives is obtained. The\nrepresentation possesses an attractive feature for applications: for all real\nvalues of the spectral parameter $\\omega$ the difference between the exact\nsolution and the approximate one (the truncated NSBF) depends on $N$ (the\ntruncation parameter) and the coefficients of the equation and does not depend\non $\\omega$. A similar result is valid when $\\omega\\in\\mathbb{C}$ belongs to a\nstrip $|Im\\omega|<C$. This feature makes the NSBF representation especially\nuseful for applications requiring computation of solutions for large intervals\nof $\\omega$. Error and decay rate estimates are obtained. An algorithm for\nsolving initial value, boundary value or spectral problems for the\nSturm-Liouville equation is developed and illustrated on a test problem.\n", " A polygon C is an intersecting polygon for a set O of objects in the plane if\nC intersects each object in O, where the polygon includes its interior. We\nstudy the problem of computing the minimum-perimeter intersecting polygon and\nthe minimum-area convex intersecting polygon for a given set O of objects. We\npresent an FPTAS for both problems for the case where O is a set of possibly\nintersecting convex polygons in the plane of total complexity n.\n Furthermore, we present an exact polynomial-time algorithm for the\nminimum-perimeter intersecting polygon for the case where O is a set of n\npossibly intersecting segments in the plane. So far, polynomial-time exact\nalgorithms were only known for the minimum perimeter intersecting polygon of\nlines or of disjoint segments.\n", " The study of Chemical Reaction Networks (CRN's) is a very active field.\nEarlier well-known results \\cite{Feinberg:def_01:87, Anderson:product_dist:10}\nidentify a topological quantity called deficiency, for any CRN, which, when\nexactly equal to zero, leads to a unique factorized steady-state for these\nnetworks. No results exist however for the steady states of non-zero-deficiency\nnetworks. In this paper, we show how to write the full moment-hierarchy for any\nnon-zero-deficiency CRN obeying mass-action kinetics, in terms of equations for\nthe factorial moments (FM). Using these, we can recursively predict values for\nlower moments from higher moments, reversing the procedure usually used to\nsolve moment hierarchies. We show, for non-trivial examples, that in this\nmanner we can predict to high accuracy, any moment of interest, for CRN's with\nnon-zero deficiency and non-factorizable steady states.\n", " We define signed dual volumes at all dimensions for circumcentric dual\nmeshes. We show that for pairwise Delaunay triangulations with mild boundary\nassumptions these signed dual volumes are positive. This allows the use of such\nDelaunay meshes for Discrete Exterior Calculus (DEC) because the discrete Hodge\nstar operator can now be correctly defined for such meshes. This operator is\ncrucial for DEC and is a diagonal matrix with the ratio of primal and dual\nvolumes along the diagonal. A correct definition requires that all entries be\npositive. DEC is a framework for numerically solving differential equations on\nmeshes and for geometry processing tasks and has had considerable impact in\ncomputer graphics and scientific computing. Our result allows the use of DEC\nwith a much larger class of meshes than was previously considered possible.\n", " In the classical facility location problem we consider a graph $G$ with fixed\nweights on the edges of $G$. The goal is then to find an optimal positioning\nfor a set of facilities on the graph with respect to some objective function.\nWe introduce a new framework for facility location problems, where the weights\non the graph edges are not fixed, but rather should be assigned. The goal is to\nfind a valid assignment for which the resulting weighted graph optimizes the\nfacility location objective function. We present algorithms for finding the\noptimal {\\em budget allocation} for the center point problem and for the median\npoint problem on trees. Our algorithms run in linear time, both for the case\nwhere a candidate vertex is given as part of the input, and for the case where\nfinding a vertex that optimizes the solution is part of the problem. We also\npresent a hardness result for the general graph case of the center point\nproblem, followed by an $O(\\log^2(n))$ approximation algorithm on graphs - with\ngeneral metric spaces.\n", " We give necessary conditions for a set E to be removable for Holder\ncontinuous quasiregular mappings in the plane. We also obtain some removability\nresults for Holder continuous mappings of finite distortion.\n", " We consider fractional differential equations of order $\\alpha \\in (0,1)$ for\nfunctions of one independent variable $t\\in (0,\\infty)$ with the\nRiemann-Liouville and Caputo-Dzhrbashyan fractional derivatives. A precise\nestimate for the order of growth of $\\alpha$-entire solutions is given. An\nanalog of the Frobenius method for systems with regular singularity is\ndeveloped. For a model example of an equation with a kind of an irregular\nsingularity, a series for a formal solution is shown to be convergent for $t>0$\n(if $\\alpha$ is an irrational number poorly approximated by rational ones) but\ndivergent in the distribution sense.\n", " What kind of reduced monomial schemes can be obtained as a Gr\\\"obner\ndegeneration of a smooth projective variety? Our conjectured answer is: only\nStanley-Reisner schemes associated to acyclic Cohen-Macaulay simplicial\ncomplexes. This would imply, in particular, that only curves of genus zero have\nsuch a degeneration. We prove this conjecture for degrevlex orders, for\nelliptic curves over real number fields, for boundaries of cross-polytopes, and\nfor leafless graphs. We discuss consequences for rational and F-rational\nsingularities of algebras with straightening laws.\n", " The threshold law for N-body fragmentation under dipole forces is formulated.\nIt emerges from the energy dependence of the normalization of the correlated\ncontinuum wave function for N fragments. It is shown that the dipole threshold\nlaw plays a key role in understanding all threshold fragmentation phenomena\nsince it links the classical threshold law for long-range Coulomb interactions\nto the statistical law for short-range interactions. Furthermore, a tunnelling\nmechanism is identified as the common feature which occurs for all three\nclasses of interactions, short-range, dipole and Coulomb.\n", " We review models and approaches for error-control in order to prevent the\nbuildup of noise when gates for digital chemical and biomolecular computing\nbased on (bio)chemical reaction processes are utilized to realize stable,\nscalable networks for information processing. Solvable rate-equation models\nillustrate several recently developed methodologies for gate-function\noptimization. We also survey future challenges and possible new research\navenues.\n", " We find necessary density conditions for Marcinkiewicz-Zygmund inequalities\nand interpolation for spaces of spherical harmonics with respect to the L^p\nnorm. Moreover, we prove that there are no complete interpolation families for\np\\neq 2.\n", " Neutrino absorption cross sections for 71Ga are calculated for all solar\nneutrino sources with standard energy spectra, and for laboratory sources of\n51Cr and 37Ar; the calculations include, where appropriate, the thermal energy\nof fusing solar ions and use improved nuclear and atomic data. The ratio, R, of\nmeasured (in GALLEX and SAGE) to calculated 51Cr capture rate is R = 0.95 +/-\n0.07 (exp)} + ^{+0.04}_{-0.03} (theory). Cross sections are also calculated for\nspecific neutrino energies chosen so that a spline fit determines accurately\nthe event rates in a gallium detector even if new physics changes the energy\nspectrum of solar neutrinos. Theoretical uncertainties are estimated for cross\nsections at specific energies and for standard neutrino energy spectra.\nStandard energy spectra are presented for pp and CNO neutrino sources in the\nappendices. Neutrino fluxes predicted by standard solar models, corrected for\ndiffusion, have been in the range 120 SNU to 141 SNU since 1968.\n", " We show that the method in recent work of Roncal, Shrivastava, and Shuin can\nbe adapted to show that certain $L^p$-improving bounds in the interior of the\nboundedness region for the bilinear spherical or triangle averaging operator\nimply sparse bounds for the corresponding lacunary maximal operator, and that\n$L^p$-improving bounds in the interior of the boundedness region for the\ncorresponding single-scale maximal operators imply sparse bounds for the\ncorrepsonding full maximal operators. More generally we show that the framework\napplies for bilinear convolutions with compactly supported finite Borel\nmeasures that satisfy appropriate $L^p$-improving and continuity estimates.\nThis shows that the method used by Roncal, Shrivastava, and Shuin can be\nadapted to obtain sparse bounds for a general class of bilinear operators that\nare not of product type, for a certain range of $L^p$ exponents.\n" ]

[ " Let $\\mathcal{C}$ be a set of $m$ by $n$ matrices over $\\mathbb{F}_q$ such\nthat the rank of $A-B$ is at least $d$ for all distinct $A,B\\in \\mathcal{C}$.\nSuppose that $m\\leqslant n$. If $\\#\\mathcal{C}= q^{n(m-d+1)}$, then\n$\\mathcal{C}$ is a maximum rank distance (MRD for short) code. Until 2016,\nthere were only two known constructions of MRD codes for arbitrary $1<d<m-1$.\nOne was found by Delsarte (1978) and Gabidulin (1985) independently, and it was\nlater generalized by Kshevetskiy and Gabidulin (2005). We often call them\n(generalized) Gabidulin codes. Another family was recently obtained by Sheekey\n(2016), and its elements are called twisted Gabidulin codes. In the same paper,\nSheekey also proposed a generalization of the twisted Gabidulin codes. However\nthe equivalence problem for it is not considered, whence it is not clear\nwhether there exist new MRD codes in this generalization. We call the members\nof this putative larger family generalized twisted Gabidulin codes. In this\npaper, we first compute the Delsarte duals and adjoint codes of them, then we\ncompletely determine the equivalence between different generalized twisted\nGabidulin codes. In particular, it can be proven that, up to equivalence,\ngeneralized Gabidulin codes and twisted Gabidulin codes are both proper subsets\nof this family.\n", " Convolutional codes are constructed, designed and analysed using row and/or\nblock structures of unit algebraic schemes. Infinite series of such codes and\nof codes with specific properties are derived. Properties are shown\nalgebraically and algebraic decoding methods are derived. For a given rate and\ngiven error-correction capability at each component, convolutional codes with\nthese specifications and with efficient decoding algorithms are constructed.\nExplicit prototype examples are given but in general large lengths and large\nerror capability are achievable. Convolutional codes with efficient decoding\nalgorithms at or near the maximum free distances attainable for the parameters\nare constructible. Unit memory convolutional codes of maximum possible free\ndistance are designed with practical algebraic decoding algorithms.\n LDPC (low density parity check) convolutional codes with efficient decoding\nschemes are constructed and analysed by the methods. Self-dual and\ndual-containing convolutional codes may also be designed by the methods;\ndual-containing codes enables the construction of quantum codes.\n", " Based on superfluid behavior of a (boson) dark matter as the light itself, a\nunified model for dark matter and quintessence is proposed. Inspired by (O'Dell\net al. 2000) which in an exciting study showed that particular configurations\nof intense off-resonant laser beams (simulate the Bose-Einstein condensate) can\ngive rise to an attractive 1/r interatomic potential, I infer electromagnetic\norigin of inertia and gravity. Time varying dark components result in time\nvarying electromagnetic fine structure constant (alpha) or the quantities given\nin its definition (even an invariant alpha would not exclude the possibility\nthat these values might be changing in ratio to one another), in either case\nvariable speed of light (VSL) with dc/dt < 0. According to (Wetterich 2003a) a\nbound R = delta_alpha(z = 0.13)/delta_alpha(z = 2) < 0.02 strongly favors\nquintessence with a time varying equation of state w = p/rho, where the value\nof (1+w) at present is substantially smaller than for z = 2 and equivalently\ncompressibility is increased in the medium and this process has significant\ninfluences on the process of structure formation. These models typically\nrequire quintessence coupling to a photon which would require large coupling\nbetween non-baryonic dark matter and the scalar field therefore consistent with\nour model, quintessence may couple to both non-baryonic dark matter and\nradiation. Wetterich (2003b) observe that Omega_quint does not affect the time\nevolution of the fundamental couplings and alpha depends only on the equation\nof state w and this fascinating result further support the common essence of\nlight and boson dark matter.\n", " We complete the classification, up to isomorphism, of the spaces of compact\noperators on C([1, gamma], l_p) spaces, 1<p< infinity. In order to do this, we\nclassify, up to isomorphism, the spaces of compact operators {\\mathcal K}(E,\nF), where E= C([1, lambda], l_p) and F=C([1,xi], l_q) for arbitrary ordinals\nlambda and xi and 1< p \\leq q< infinity.\n More precisely, we prove that it is relatively consistent with ZFC that for\nany infinite ordinals lambda, mu, xi and eta the following statements are\nequivalent:\n (a) {\\mathcal K}(C([1, lambda], l_p), C([1, xi], l_q)) is isomorphic to\n{\\mathcal K}(C([1, mu], l_p), C([1, eta], l_q)) .\n (b) lambda and mu have the same cardinality and C([1,xi]) is isomorphic to\nC([1, eta]) or there exists an uncountable regular ordinal alpha and 1 \\leq m,\nn < omega such that C([1, xi]) is isomorphic to C([1, alpha m]) and C([1,eta])\nis isomorphic to C([1, alpha n]).\n Moreover, in ZFC, if lambda and mu are finite ordinals and xi and eta are\ninfinite ordinals then the statements (a) and (b') are equivalent.\n (b') C([1,xi]) is isomorphic to C([1, eta]) or there exists an uncountable\nregular ordinal alpha and 1 \\leq m, n \\leq omega such that C([1, xi]) is\nisomorphic to C([1, alpha m]) and C([1,eta]) is isomorphic to C([1, alpha n]).\n", " Let $L$ be a set of $n$ lines in $R^3$ that is contained, when represented as\npoints in the four-dimensional Pl\\\"ucker space of lines in $R^3$, in an\nirreducible variety $T$ of constant degree which is \\emph{non-degenerate} with\nrespect to $L$ (see below). We show:\n \\medskip \\noindent{\\bf (1)} If $T$ is two-dimensional, the number of $r$-rich\npoints (points incident to at least $r$ lines of $L$) is\n$O(n^{4/3+\\epsilon}/r^2)$, for $r \\ge 3$ and for any $\\epsilon>0$, and, if at\nmost $n^{1/3}$ lines of $L$ lie on any common regulus, there are at most\n$O(n^{4/3+\\epsilon})$ $2$-rich points. For $r$ larger than some sufficiently\nlarge constant, the number of $r$-rich points is also $O(n/r)$.\n As an application, we deduce (with an $\\epsilon$-loss in the exponent) the\nbound obtained by Pach and de Zeeuw (2107) on the number of distinct distances\ndetermined by $n$ points on an irreducible algebraic curve of constant degree\nin the plane that is not a line nor a circle.\n \\medskip \\noindent{\\bf (2)} If $T$ is two-dimensional, the number of\nincidences between $L$ and a set of $m$ points in $R^3$ is $O(m+n)$.\n \\medskip \\noindent{\\bf (3)} If $T$ is three-dimensional and nonlinear, the\nnumber of incidences between $L$ and a set of $m$ points in $R^3$ is\n$O\\left(m^{3/5}n^{3/5} + (m^{11/15}n^{2/5} + m^{1/3}n^{2/3})s^{1/3} + m + n\n\\right)$, provided that no plane contains more than $s$ of the points. When $s\n= O(\\min\\{n^{3/5}/m^{2/5}, m^{1/2}\\})$, the bound becomes\n$O(m^{3/5}n^{3/5}+m+n)$.\n As an application, we prove that the number of incidences between $m$ points\nand $n$ lines in $R^4$ contained in a quadratic hypersurface (which does not\ncontain a hyperplane) is $O(m^{3/5}n^{3/5} + m + n)$.\n The proofs use, in addition to various tools from algebraic geometry, recent\nbounds on the number of incidences between points and algebraic curves in the\nplane.\n", " Starting from the meaning of the conjugate of a complex Hilbert space,\nincluding a related application of the theorem of Fr\\'{e}chet-Riesz (by which\nan analysis of semilinear operators can be reduced to - linear - operator\ntheory) to a revisit of applications of nuclear and absolutely $p$-summing\noperators in algebraic quantum field theory in the sense of Araki, Haag and\nKastler ($p=2$) and more recently in the framework of general probabilistic\nspaces ($p=1$), we will outline that Banach operator ideals in the sense of\nPietsch, or equivalently tensor products of Banach spaces in the sense of\nGrothendieck are even lurking in the foundations and philosophy of quantum\nphysics and quantum information theory. In particular, we concentrate on their\nimportance in algebraic quantum field theory. In doing so, we establish a\ncanonical isometric isomorphism between the Hilbert spaces $H\\otimes_2 (K\n\\otimes_2 L)$ and $(H \\otimes_2 K) \\otimes_2 L$ (Theorem 3.8) and revisit the\nrole of trace class operators.\n A few applications are specified, including the appropriateness of the class\nof Hilbert-Schmidt operators and an implied Banach operator ideal\nrepresentation of the tensor product of two complex Hilbert spaces $H \\otimes_2\nK$ (Proposition 3.4) and a purely linear algebraic description of the quantum\nteleportation process (Example 3.10).\n", " In this paper we present new data structures for two extensively studied\nvariants of the orthogonal range searching problem.\n First, we describe a data structure that supports two-dimensional orthogonal\nrange minima queries in $O(n)$ space and $O(\\log^{\\varepsilon} n)$ time, where\n$n$ is the number of points in the data structure and $\\varepsilon$ is an\narbitrarily small positive constant. Previously known linear-space solutions\nfor this problem require $O(\\log^{1+\\varepsilon} n)$ (Chazelle, 1988) or\n$O(\\log n\\log \\log n)$ time (Farzan et al., 2012). A modification of our data\nstructure uses space $O(n\\log \\log n)$ and supports range minima queries in\ntime $O(\\log \\log n)$. Both results can be extended to support\nthree-dimensional five-sided reporting queries.\n Next, we turn to the four-dimensional orthogonal range reporting problem and\npresent a data structure that answers queries in optimal $O(\\log n/\\log \\log n\n+ k)$ time, where $k$ is the number of points in the answer. This is the first\ndata structure that achieves the optimal query time for this problem.\n Our results are obtained by exploiting the properties of three-dimensional\nshallow cuttings.\n", " Given a metric space $(X,d)$, the wobbling group of $X$ is the group of\nbijections $g:X\\rightarrow X$ satisfying $\\sup\\limits_{x\\in X}\nd(g(x),x)<\\infty$. We study algebraic and analytic properties of $W(X)$ in\nrelation with the metric space structure of $X$, such as amenability of the\naction of the lamplighter group $ \\bigoplus_{X} \\mathbf Z/2\\mathbf Z \\rtimes\nW(X)$ on $\\bigoplus_{X} \\mathbf Z/2\\mathbf Z$ and property (T).\n", " In the present paper we introduce and study finite point subsets of a special\nkind, called optimum distributions, in the n-dimensional unit cube. Such\ndistributions are closely related with known (delta,s,n)-nets of low\ndiscrepancy. It turns out that optimum distributions have a rich combinatorial\nstructure. Namely, we show that optimum distributions can be characterized\ncompletely as maximum distance separable codes with respect to a non-Hamming\nmetric. Weight spectra of such codes can be evaluated precisely. We also\nconsider linear codes and distributions and study their general properties\nincluding the duality with respect to a suitable inner product. The\ncorresponding generalized MacWilliams identities for weight enumerators are\nbriefly discussed. Broad classes of linear maximum distance separable codes and\nlinear optimum distributions are explicitely constructed in the paper by the\nHermite interpolations over finite fields.\n", " We present a short review of the current quasar (QSO) absorption line\nconstraints on possible variation of the fine structure constant, alpha =\ne^2/(hbar*c). Particular attention is paid to recent optical Keck/HIRES spectra\nof 49 absorption systems which indicate a smaller alpha in the past (Murphy et\nal. 2001, Webb et al. 2001). Here we present new preliminary results from 128\nabsorption systems: da/a = (-0.57 +/- 0.10) x 10^{-5} over the redshift range\n0.2 < z < 3.7, in agreement with the previous results. Known potential\nsystematic errors cannot explain these results. We compare them with strong\n`local' constraints and discuss other (radio and millimetre-wave) QSO\nabsorption line constraints on variations in alpha^2 * g_p and alpha^2 * g_p *\nm_e/m_p (g_p is the proton g-factor and m_e/m_p is the electron/proton mass\nratio). Finally, we discuss future efforts to rule out or confirm the current\n5.7 sigma optical detection.\n", " In the light of recent neutrino oscillation and non-oscillation data, we\nrevisit the phenomenological constraints applicable to three observables\nsensitive to absolute neutrino masses: The effective neutrino mass in single\nbeta decay (m_beta); the effective Majorana neutrino mass in neutrinoless\ndouble beta decay (m_2beta); and the sum of neutrino masses in cosmology\n(Sigma). In particular, we include the constraints coming from the first Main\nInjector Neutrino Oscillation Search (MINOS) data and from the Wilkinson\nMicrowave Anisotropy Probe (WMAP) three-year (3y) data, as well as other\nrelevant cosmological data and priors. We find that the largest neutrino\nsquared mass difference is determined with a 15% accuracy (at 2-sigma) after\nadding MINOS to world data. We also find upper bounds on the sum of neutrino\nmasses Sigma ranging from ~2 eV (WMAP-3y data only) to ~0.2 eV (all\ncosmological data) at 2-sigma, in agreement with previous studies. In addition,\nwe discuss the connection of such bounds with those placed on the matter power\nspectrum normalization parameter sigma_8. We show how the partial degeneracy\nbetween Sigma and sigma_8 in WMAP-3y data is broken by adding further\ncosmological data, and how the overall preference of such data for relatively\nhigh values of sigma_8 pushes the upper bound of Sigma in the sub-eV range.\nFinally, for various combination of data sets, we revisit the (in)compatibility\nbetween current Sigma and m_2beta constraints (and claims), and derive\nquantitative predictions for future single and double beta decay experiments.\n", " The equivariant motivic Chern class of a Schubert cell in a `complete' flag\nmanifold $X=G/B$ is an element in the equivariant K theory ring of $X$ to which\none adjoins a formal parameter $y$. In this paper we prove several `folklore\nresults' about the motivic Chern classes, including finding specializations at\n$y=-1$ and $y=0$; the coefficient of the top power of $y$; how to obtain\nChern-Schwartz-MacPherson (CSM) classes as leading terms of motivic classes;\ndivisibility properties of the Schubert expansion of motivic Chern classes. We\ncollect several conjectures about the positivity, unimodality, and log\nconcavity of CSM and motivic Chern classes of Schubert cells, including a\nconjectural positivity of structure constants of the multiplication of\nPoincar\\'e duals of CSM classes. In addition, we prove a `star duality' for the\nmotivic Chern classes. We utilize the motivic Chern transformation to define\ntwo equivariant variants of the Hirzebruch transformation, which appear\nnaturally in the Grothendieck-Hirzebruch-Riemann-Roch formalism. We utilize the\nDemazure-Lusztig recursions from the motivic Chern class theory to find similar\nrecursions giving the Hirzebruch classes of Schubert cells, their Poincar{\\'e}\nduals, and their Segre versions. We explain the functoriality properties needed\nto extend the results to `partial' flag manifolds $G/P$.\n", " Let G=(V,E) be a graph. A set S is independent if no two vertices from S are\nadjacent. The independence number alpha(G) is the cardinality of a maximum\nindependent set, and mu(G) is the size of a maximum matching. The number\nid_{c}(G)=max{|I|-|N(I)|:I is an independent set} is called the critical\nindependence difference of G, and A is critical if |A|-|N(A)|=id_{c}(G). We\ndefine core(G) as the intersection of all maximum independent sets, and\nker(G)as the intersection of all critical independent sets. In this paper we\nprove that if a graph G is non-quasi-regularizable (i.e., there exists some\nindependent set A, such that |A|>|N(A)|), then: ker(G) is a subset of core(G),\nand |ker(G)|> id_{c}(G) >= alpha(G)-mu(G) > 0.\n", " In this study, we define new paranormed sequence spaces by the sequences of\nFibonacci numbers. Furthermore, we compute the $\\alpha-,\\beta-$ and $\\gamma-$\nduals and obtain bases for these sequence spaces. Besides this, we characterize\nthe matrix transformations from the new paranormed sequence spaces to the\nMaddox's spaces $c_{0}(q),c(q),\\ell(q)$ and $\\ell_{\\infty}(q)$.\n", " As for the Fourier transforms of positive and integrable functions supported\nin the unit interval, we make a list of improvements for P\\'olya's results on\nthe distribution of their positive zeros and give new sufficient conditions\nunder which those zeros are simple and regularly distributed. As an\napplication, we take the two-parameter family of beta probability density\nfunctions defined by \\begin{equation*} f(t)= \\frac{1}{B(\\alpha, \\beta)}\\,\n(1-t)^{\\alpha-1} t^{\\beta-1},\\quad 0<t<1, \\end{equation*} where\n$\\,\\alpha>0,\\,\\beta>0,$ and specify the distribution of zeros of the associated\nFourier transforms for some region of $(\\alpha, \\beta)$ in the first quadrant\nwhich turns out to be much larger than the region where P\\'olya's results are\napplicable.\n", " The alpha complex is a fundamental data structure from computational\ngeometry, which encodes the topological type of a union of balls $B(x; r)\n\\subset \\mathbb{R}^m$ for $x\\in S$, including a weighted version that allows\nfor varying radii. It consists of the collection of \"simplices\" $\\sigma =\n\\{x_0, ..., x_k \\} \\subset S$, which correspond to nomempty $(k + 1)$-fold\nintersections of cells in a radius-restricted version of the Voronoi diagram.\nExisting algorithms for computing the alpha complex require that the points\nreside in low dimension because they begin by computing the entire Delaunay\ncomplex, which rapidly becomes intractable, even when the alpha complex is of a\nreasonable size. This paper presents a method for computing the alpha complex\nwithout computing the full Delaunay triangulation by applying Lagrangian\nduality, specifically an algorithm based on dual quadratic programming that\nseeks to rule simplices out rather than ruling them in.\n", " The $k$-token graph $F_k(G)$ of a graph $G$ is the graph whose vertices are\nthe $k$-subsets of vertices from $G$, two of which being adjacent whenever\ntheir symmetric difference is a pair of adjacent vertices in $G$. It is a known\nresult that the algebraic connectivity (or second Laplacian eigenvalue) of\n$F_k(G)$ equals the algebraic connectivity of $G$.\n In this paper, we first give results that relate the algebraic connectivities\nof a token graph and the same graph after removing a vertex. Then, we prove the\nresult on the algebraic connectivity of 2-token graphs for two infinite\nfamilies: the odd graphs $O_r$ for all $r$, and the multipartite complete\ngraphs $K_{n_1,n_2,\\ldots,n_r}$ for all $n_1,n_2,\\ldots,n_r$ In the case of\ncycles, we present a new method that allows us to compute the whole spectrum of\n$F_2(C_n)$. This method also allows us to obtain closed formulas that give\nasymptotically exact approximations for most of the eigenvalues of\n$F_2(\\textit{}C_n)$.\n", " We present a major revamp of the point-location data structure for general\ntwo-dimensional subdivisions via randomized incremental construction,\nimplemented in CGAL, the Computational Geometry Algorithms Library. We can now\nguarantee that the constructed directed acyclic graph G is of linear size and\nprovides logarithmic query time. Via the construction of the Voronoi diagram\nfor a given point set S of size n, this also enables nearest-neighbor queries\nin guaranteed O(log n) time. Another major innovation is the support of general\nunbounded subdivisions as well as subdivisions of two-dimensional parametric\nsurfaces such as spheres, tori, cylinders. The implementation is exact,\ncomplete, and general, i.e., it can also handle non-linear subdivisions. Like\nthe previous version, the data structure supports modifications of the\nsubdivision, such as insertions and deletions of edges, after the initial\npreprocessing. A major challenge is to retain the expected O(n log n)\npreprocessing time while providing the above (deterministic) space and\nquery-time guarantees. We describe an efficient preprocessing algorithm, which\nexplicitly verifies the length L of the longest query path in O(n log n) time.\nHowever, instead of using L, our implementation is based on the depth D of G.\nAlthough we prove that the worst case ratio of D and L is Theta(n/log n), we\nconjecture, based on our experimental results, that this solution achieves\nexpected O(n log n) preprocessing time.\n", " The Hilbert number $H(n)$ is defined as the maximum number of limit cycles of\na planar autonomous system of ordinary differential equations (ODEs) with\nright-hand sides containing polynomials of degree at most $n \\in {\\mathbb N}$.\nThe dynamics of chemical reaction systems with two chemical species can be\n(under mass-action kinetics) described by such planar autonomous ODEs, where\n$n$ is equal to the maximum order of the chemical reactions in the system.\nGeneralizations of the Hilbert number $H(n)$ to three different classes of\nchemical reaction networks are investigated: (i) chemical systems with\nreactions up to the $n$-th order; (ii) systems with up to $n$-molecular\nchemical reactions; and (iii) weakly reversible chemical reaction networks. In\neach case (i), (ii) and (iii), the question on the number of limit cycles is\nconsidered. Lower bounds on the generalized Hilbert numbers are provided for\nboth algebraic and non-algebraic limit cycles. Furthermore, given a general\nalgebraic curve $h(x,y)=0$ of degree $n_h \\in {\\mathbb N}$ and containing one\nor more ovals in the positive quadrant, a chemical system is constructed which\nhas the oval(s) as its stable algebraic limit cycle(s). The ODEs describing the\ndynamics of the constructed chemical system contain polynomials of degree at\nmost $n=2\\,n_h+1.$ Considering $n_h \\ge 4,$ the algebraic curve $h(x,y)=0$ can\ncontain multiple closed components with the maximum number of ovals given by\nHarnack's curve theorem as $1+(n_h-1)(n_h-2)/2$, which is equal to 4 for\n$n_h=4.$ Algebraic curve $h(x,y)=0$ with $n_h=4$ and the maximum number of four\novals is used to construct a chemical system which has four stable algebraic\nlimit cycles.\n", " In this work, we study matroids over a domain and several classical\ncombinatorial and algebraic invariants related. We define their\nGrothendieck-Tutte polynomial $T_{\\mathcal{M}}(x,y)$, extending the definition\ngiven by Fink and Moci in 2016, and we show that such polynomial has the\nclassical deletion-contraction property.\n Moreover, we study the set of independencies for a realizable matroid over a\ndomain, generalizing the definition of \\emph{poset of torsions}\n$Gr(\\mathcal{M})$ given by the second author in 2017. This is a union of\nidentical simplicial posets as for (quasi-)arithmetic matroids. The new notions\nharmonize naturally through the face module $N_\\mathcal{M}$ of the matroid over\na domain.\n Whenever $Gr(\\mathcal{M})$ is a finite poset, the Hilbert series\n$N_\\mathcal{M}(t)$ of its face module is a specialization of the Tutte\npolynomial $T_{\\mathcal{M}}(x,y)$. Further, for arrangements of codimension-one\nabelian subvarities of an elliptic curve admitting complex multiplication, we\nextend certain results of Bibby and we provide an algebraic interpretation of\nthe elliptic Tutte polynomial.\n", " We consider Cauchy type integrals $I(t)={1\\over 2\\pi i}\\int_{\\gamma}\n{g(z)dz\\over z-t}$ with $g(z)$ an algebraic function. The main goal is to give\nconstructive (at least, in principle) conditions for $I(t)$ to be an algebraic\nfunction, a rational function, and ultimately an identical zero near infinity.\nThis is done by relating the Monodromy group of the algebraic function $g$, the\ngeometry of the integration curve $\\gamma$, and the analytic properties of the\nCauchy type integrals. The motivation for the study of these conditions is\nprovided by the fact that certain Cauchy type integrals of algebraic functions\nappear in the infinitesimal versions of two classical open questions in\nAnalytic Theory of Differential Equations: the Poincar\\'e Center-Focus problem\nand the second part of the Hilbert 16-th problem.\n", " Neural codes serve as a language for neurons in the brain. Convex codes,\nwhich arise from the pattern of intersections of convex sets in Euclidean\nspace, are of particular relevance to neuroscience. Not every code is convex,\nhowever, and the combinatorial properties of a code that determine its\nconvexity are still poorly understood. Here we find that a code that can be\nrealized by a collection of open convex sets may or may not be realizable by\nclosed convex sets, and vice versa, establishing that open convex and closed\nconvex codes are distinct classes. We also prove that max intersection-complete\ncodes (i.e. codes that contain all intersections of maximal codewords) are both\nopen convex and closed convex, and provide an upper bound for their minimal\nembedding dimension. Finally, we show that the addition of non-maximal\ncodewords to an open convex code preserves convexity.\n", " We construct bijections giving three \"codes\" for trees. These codes follow\nnaturally from the Matrix Tree Theorem of Tutte and have many advantages over\nthe one produced by Prufer in 1918. One algorithm gives explicitly a bijection\nthat is implicit in Orlin's manipulatorial proof of Cayley's formula (the\nformula was actually found first by Borchardt). Another is based on a proof of\nKnuth. The third is an implementation of Joyal's pseudo-bijective proof of the\nformula, and is equivalent to one previously found by Egecioglu and Remmel. In\neach case, we have at least two algorithms, one of which involves hands-on\nmanipulations of the tree while the other involves a combinatorial and linear\nalgebraic manipulation of a matrix.\n", " We investigate the bounds on algebraic connectivity of graphs subject to\nconstraints on the number of edges, vertices, and topology. We show that the\nalgebraic connectivity for any tree on $n$ vertices and with maximum degree $d$\nis bounded above by $2(d-2) \\frac{1}{n}+O(\\frac{\\ln n}{n^{2}}) .$ We then\ninvestigate upper bounds on algebraic connectivity for cubic graphs. We show\nthat algebraic connectivity of a cubic graph of girth $g$ is bounded above by\n$3-2^{3/2}\\cos(\\pi/\\lfloor g/2\\rfloor) ,$ which is an improvement over the\nbound found by Nilli [A. Nilli, Electron. J. Combin., 11(9), 2004]. Finally, we\npropose several conjectures and open questions.\n", " The set of permutations on a finite set can be given the lattice structure\nknown as the weak Bruhat order. This lattice structure is generalized to the\nset of words on a fixed alphabet $\\Sigma$ = {x,y,z,...}, where each letter has\na fixed number of occurrences. These lattices are known as multinomial lattices\nand, when card($\\Sigma$) = 2, as lattices of lattice paths. By interpreting the\nletters x, y, z, . . . as axes, these words can be interpreted as discrete\nincreasing paths on a grid of a d-dimensional cube, with d = card($\\Sigma$).We\nshow how to extend this ordering to images of continuous monotone functions\nfrom the unit interval to a d-dimensional cube and prove that this ordering is\na lattice, denoted by L(I^d). This construction relies on a few algebraic\nproperties of the quantale of join-continuous functions from the unit interval\nof the reals to itself: it is cyclic $\\star$-autonomous and it satisfies the\nmix rule.We investigate structural properties of these lattices, which are\nself-dual and not distributive. We characterize join-irreducible elements and\nshow that these lattices are generated under infinite joins from their\njoin-irreducible elements, they have no completely join-irreducible elements\nnor compact elements. We study then embeddings of the d-dimensional multinomial\nlattices into L(I^d). We show that these embeddings arise functorially from\nsubdivisions of the unit interval and observe that L(I^d) is the\nDedekind-MacNeille completion of the colimit of these embeddings. Yet, if we\nrestrict to embeddings that take rational values and if d > 2, then every\nelement of L(I^d) is only a join of meets of elements from the colimit of these\nembeddings.\n", " An external memory data structure is presented for maintaining a dynamic set\nof $N$ two-dimensional points under the insertion and deletion of points, and\nsupporting 3-sided range reporting queries and top-$k$ queries, where top-$k$\nqueries report the $k$~points with highest $y$-value within a given $x$-range.\nFor any constant $0<\\varepsilon\\leq \\frac{1}{2}$, a data structure is\nconstructed that supports updates in amortized $O(\\frac{1}{\\varepsilon\nB^{1-\\varepsilon}}\\log_B N)$ IOs and queries in amortized\n$O(\\frac{1}{\\varepsilon}\\log_B N+K/B)$ IOs, where $B$ is the external memory\nblock size, and $K$ is the size of the output to the query (for top-$k$ queries\n$K$ is the minimum of $k$ and the number of points in the query interval). The\ndata structure uses linear space. The update bound is a significant factor\n$B^{1-\\varepsilon}$ improvement over the previous best update bounds for the\ntwo query problems, while staying within the same query and space bounds.\n", " Using the ideas of E.I. Gordon we present and farther advance an approach,\nbased on nonstandard analysis, to simultaneous approximations of locally\ncompact abelian groups and their duals by (hyper)finite abelian groups, as well\nas to approximations of various types of Fourier transforms on them by the\ndiscrete Fourier transform. Combining some methods of nonstandard analysis and\nadditive combinatorics we prove the three Gordon's Conjectures which were open\nsince 1991 and are crucial both in the formulations and proofs of the LCA\ngroups and Fourier transform approximation theorems.\n", " Universe structure emerges in the unreduced, complex-dynamic interaction\nprocess with the simplest initial configuration (two attracting homogeneous\nfields, quant-ph/9902015). The unreduced interaction analysis gives\nintrinsically creative cosmology, describing the real, explicitly emerging\nworld structure with dynamic randomness on each scale. Without imposing any\npostulates or entities, we obtain physically real space, time, elementary\nparticles with their detailed structure and intrinsic properties, causally\ncomplete and unified version of quantum and relativistic behaviour, the origin\nand number of naturally unified fundamental forces, and classical behaviour\nemergence in closed systems (gr-qc/9906077). Main problems of standard\ncosmology and astrophysics are consistently solved in this extended picture\n(without introduction of any additional entities), including those of quantum\ncosmology and gravity, entropy growth and time, \"hierarchy\" of elementary\nparticles, \"anthropic\" difficulties, space-time flatness, and \"missing\"\n(\"dark\") mass and energy. The observed universe structure and laws can be\npresented as manifestations of the universal symmetry (conservation) of\ncomplexity providing the unified, irregular, but exact (never \"broken\") Order\nof the World (physics/0404006).\n", " Generic Painlev\\'e VI tau function \\tau(t) can be interpreted as four-point\ncorrelator of primary fields of arbitrary dimensions in 2D CFT with c=1. Using\nAGT combinatorial representation of conformal blocks and determining the\ncorresponding structure constants, we obtain full and completely explicit\nexpansion of \\tau(t) near the singular points. After a check of this expansion,\nwe discuss examples of conformal blocks arising from Riccati, Picard, Chazy and\nalgebraic solutions of Painlev\\'e VI.\n", " We extend (and somewhat simplify) the algebraic proof technique of Guth and\nKatz \\cite{GK}, to obtain several sharp bounds on the number of incidences\nbetween lines and points in three dimensions. Specifically, we show: (i) The\nmaximum possible number of incidences between $n$ lines in $\\reals^3$ and $m$\nof their joints (points incident to at least three non-coplanar lines) is\n$\\Theta(m^{1/3}n)$ for $m\\ge n$, and $\\Theta(m^{2/3}n^{2/3}+m+n)$ for $m\\le n$.\n(ii) In particular, the number of such incidences cannot exceed $O(n^{3/2})$.\n(iii) The bound in (i) also holds for incidences between $n$ lines and $m$\narbitrary points (not necessarily joints), provided that no plane contains more\nthan O(n) points and each point is incident to at least three lines. As a\npreliminary step, we give a simpler proof of (an extension of) the bound\n$O(n^{3/2})$, established by Guth and Katz, on the number of joints in a set of\n$n$ lines in $\\reals^3$. We also present some further extensions of these\nbounds, and give a proof of Bourgain's conjecture on incidences between points\nand lines in 3-space, which constitutes a simpler alternative to the proof of\n\\cite{GK}.\n", " We construct a complete list of O(q^4) terms directly within Heavy Baryon\nChiral Perturbation Theory (HBChPT) in the absence of external fields and\nassuming isospin symmetry. In addition to a phase rule recently developed\n(NPA646, 343, '99), a variety of algebraic identities and reparameterization\ninvariance, are used to ensure linear independence of the terms and their low\nenergy coupling constants. We first construct O(q^4) terms terms for off-shell\nnucleons, and then perform the on-shell reduction, again within HBChPT. We\ndiscuss an application of the O(q^4) terms to the evaluation of the O(q^4)\noperator insertions in the \"contact graph\" of pion double charge exchange.\n", " Investigation of the nucleon's excited states has always become an important\nresearch topic because of the rich information they provide. Since their first\nobservation, dating back about 70 years, the investigation of their various\nparameters contributed both to the development of the quark model and better\nunderstanding of the QCD as the theory of strong interaction. Their\ninvestigation still has importance. The researches conducted on the nucleon\nexcited states are helpful to probe missing resonances predicted by the quark\nmodel but not observed yet. With this motivation, we study the low lying\nnucleon resonance with $I(J^P)=\\frac{1}{2}(\\frac{3}{2}^{-})$ and its\ncorresponding orbital and radial excitations with\n$I(J^P)=\\frac{1}{2}(\\frac{3}{2}^+)$ and $I(J^P)=\\frac{1}{2}(\\frac{3}{2}^-)$,\nrespectively. Using the QCD sum rule method, we calculate the masses and pole\nresidues of these states. The obtained mass results are consistent with the\nmass ranges presented in PDG for the resonances $N^*(1520)(3/2^-)$,\n$N^*(1700)(3/2^-)$, and $N^*(1720)(3/2^+)$. The results of masses and residues\nof these states may be used as input parameters to calculate various quantities\nrelated to their electromagnetic, weak and strong interactions with other\nparticles with the aim of getting more information on their nature and\nstructure.\n", " We consider the generalised Beta function introduced by Chaudhry {\\it et\nal.\\/} [J. Comp. Appl. Math. {\\bf 78} (1997) 19--32] defined by\n\\[B(x,y;p)=\\int_0^1 t^{x-1} (1-t)^{y-1} \\exp\n\\left[\\frac{-p}{4t(1-t)}\\right]\\,dt,\\] where $\\Re (p)>0$ and the parameters $x$\nand $y$ are arbitrary complex numbers. The asymptotic behaviour of $B(x,y;p)$\nis obtained when (i) $p$ large, with $x$ and $y$ fixed, (ii) $x$ and $p$ large,\n(iii) $x$, $y$ and $p$ large and (iv) either $x$ or $y$ large, with $p$ finite.\nNumerical results are given to illustrate the accuracy of the formulas\nobtained.\n", " Previous work on convexity of neural codes has produced codes that are\nopen-convex but not closed-convex -- or vice-versa. However, why a code is one\nbut not the other, and how to detect such discrepancies are open questions. We\ntackle these questions in two ways. First, we investigate the concept of\ndegeneracy introduced by Cruz et al., and extend their results to show that\nnondegeneracy precisely captures the situation when taking closures or\ninteriors of open or closed realizations, respectively, yields another\nrealization of the code. Second, we give the first general criteria for\nprecluding a code from being closed-convex (without ruling out open-convexity),\nunifying ad-hoc geometric arguments in prior works. One criterion is built on a\nphenomenon we call a rigid structure, while the other can be stated\nalgebraically, in terms of the neural ideal of the code. These results\ncomplement existing criteria having the opposite purpose: precluding\nopen-convexity but not closed-convexity. Finally, we show that a family of\ncodes shown by Jeffs to be not open-convex is in fact closed-convex and\nrealizable in dimension two.\n", " Let vphi:C rightarrow C be a bilipschitz map. We prove that if E\\subset\\C is\ncompact, and gamma(E), alpha(E) stand for its analytic and continuous analytic\ncapacity respectively, then C^{-1}\\gamma(E)\\leq \\gamma(\\vphi(E)) \\leq\nC\\gamma(E) and C^{-1}\\alpha(E)\\leq \\alpha(\\vphi(E)) \\leq C\\alpha(E), where C\ndepends only on the bilipschitz constant of vphi. Further, we show that if mu\nis a Radon measure on C and the Cauchy transform is bounded on L^2(\\mu), then\nthe Cauchy transform is also bounded on L^2(\\vphi_\\sharp\\mu), where\nvphi_\\sharp\\mu is the image measure of mu by vphi. To obtain these results, we\nestimate the curvature of vphi_\\sharp\\mu by means of a corona type\ndecomposition.\n", " We study the problem of constructing a data structure that can store a\ntwo-dimensional polygonal curve $P$, such that for any query segment\n$\\overline{ab}$ one can efficiently compute the Fr\\'{e}chet distance between\n$P$ and $\\overline{ab}$. First we present a data structure of size $O(n \\log\nn)$ that can compute the Fr\\'{e}chet distance between $P$ and a horizontal\nquery segment $\\overline{ab}$ in $O(\\log n)$ time, where $n$ is the number of\nvertices of $P$. In comparison to prior work, this significantly reduces the\nrequired space. We extend the type of queries allowed, as we allow a query to\nbe a horizontal segment $\\overline{ab}$ together with two points $s, t \\in P$\n(not necessarily vertices), and ask for the Fr\\'{e}chet distance between\n$\\overline{ab}$ and the curve of $P$ in between $s$ and $t$. Using\n$O(n\\log^2n)$ storage, such queries take $O(\\log^3 n)$ time, simplifying and\nsignificantly improving previous results. We then generalize our results to\nquery segments of arbitrary orientation. We present an\n$O(nk^{3+\\varepsilon}+n^2)$ size data structure, where $k \\in [1..n]$ is a\nparameter the user can choose, and $\\varepsilon > 0$ is an arbitrarily small\nconstant, such that given any segment $\\overline{ab}$ and two points $s, t \\in\nP$ we can compute the Fr\\'{e}chet distance between $\\overline{ab}$ and the\ncurve of $P$ in between $s$ and $t$ in $O((n/k)\\log^2n+\\log^4 n)$ time. This is\nthe first result that allows efficient exact Fr\\'{e}chet distance queries for\narbitrarily oriented segments.\n We also present two applications of our data structure: we show that we can\ncompute a local $\\delta$-simplification (with respect to the Fr\\'{e}chet\ndistance) of a polygonal curve in $O(n^{5/2+\\varepsilon})$ time, and that we\ncan efficiently find a translation of an arbitrary query segment\n$\\overline{ab}$ that minimizes the Fr\\'{e}chet distance with respect to a\nsubcurve of $P$.\n", " In this article we present a class of codes with few weights arising from\nspecial type of linear sets. We explicitly show the weights of such codes,\ntheir weight enumerator and possible choices for their generator matrices. In\nparticular, our construction yields also to linear codes with three weights\nand, in some cases, to almost MDS codes. The interest for these codes relies on\ntheir applications to authentication codes and secret schemes, and their\nconnections with further objects such as association schemes and graphs.\n", " Let $M$ and $N$ be fixed non-negative integer numbers and let $\\pi_N$ be a\npolynomial of degree $N$. Suppose that $(P_n)_{n\\geq0}$ and $(Q_n)_{n\\geq0}$\nare two orthogonal polynomial sequences such that %their derivatives of orders\n$k$ and $m$ (respectively) satisfy the structure relation $$\n\\pi_N(x)\\,P_{n+m}^{(m)}(x)= \\sum_{j=n-M}^{n+N}r_{n,j}Q_{j+k}^{(k)}(x)\\quad\n(n=0,1,\\ldots)\\,, $$ where $r_{n,j}$ are complex number independent of $x$. It\nis shown that under natural constraints, $(P_n)_{n\\geq0}$ and $(Q_n)_{n\\geq0}$\nare semiclassical orthogonal polynomial sequences. Moreover, their\ncorresponding moment linear functionals are related by a rational modification\nin the distributional sense. This leads to the concept of $\\pi_N-$coherent pair\nwith index $M$ and order $(m,k)$.\n", " The 0-instanton solution of Painlev\\'e I is a sequence $(u_{n,0})$ of complex\nnumbers which appears universally in many enumerative problems in algebraic\ngeometry, graph theory, matrix models and 2-dimensional quantum gravity. The\nasymptotics of the 0-instanton $(u_{n,0})$ for large $n$ were obtained by the\nthird author using the Riemann-Hilbert approach. For $k=0,1,2,...$, the\n$k$-instanton solution of Painlev\\'e I is a doubly-indexed sequence $(u_{n,k})$\nof complex numbers that satisfies an explicit quadratic non-linear recursion\nrelation. The goal of the paper is three-fold: (a) to compute the asymptotics\nof the 1-instanton sequence $(u_{n,1})$ to all orders in $1/n$ by using the\nRiemann-Hilbert method, (b) to present formulas for the asymptotics of\n$(u_{n,k})$ for fixed $k$ and to all orders in $1/n$ using resurgent analysis,\nand (c) to confirm numerically the predictions of resurgent analysis. We point\nout that the instanton solutions display a new type of Stokes behavior, induced\nfrom the tritronqu\\'ee Painlev\\'e transcendents, and which we call the induced\nStokes phenomenon. The asymptotics of the 2-instanton and beyond exhibits new\nphenomena not seen in 0 and 1-instantons, and their enumerative context is at\npresent unknown.\n", " We consider computing the Riemann zeta function $\\zeta(s)$ and Dirichlet\n$L$-functions $L(s,\\chi)$ to $p$-bit accuracy for large $p$. Using the\napproximate functional equation together with asymptotically fast computation\nof the incomplete gamma function, we observe that $p^{3/2+o(1)}$ bit complexity\ncan be achieved if $s$ is an algebraic number of fixed degree and with\nalgebraic height bounded by $O(p)$. This is an improvement over the\n$p^{2+o(1)}$ complexity of previously published algorithms and yields, among\nother things, $p^{3/2+o(1)}$ complexity algorithms for Stieltjes constants and\n$n^{3/2+o(1)}$ complexity algorithms for computing the $n$th Bernoulli number\nor the $n$th Euler number exactly.\n", " Codes in finite projective spaces equipped with the subspace distance have\nbeen proposed for error control in random linear network coding. The resulting\nso-called \\emph{Main Problem of Subspace Coding} is to determine the maximum\nsize $A_q(v,d)$ of a code in $\\operatorname{PG}(v-1,\\mathbb{F}_q)$ with minimum\nsubspace distance $d$. Here we completely resolve this problem for $d\\ge v-1$.\nFor $d=v-2$ we present some improved bounds and determine $A_q(5,3)=2q^3+2$\n(all $q$), $A_2(7,5)=34$. We also provide an exposition of the known\ndetermination of $A_q(v,2)$, and a table with exact results and bounds for the\nnumbers $A_2(v,d)$, $v\\leq 7$.\n", " Zonotopal algebra interweaves algebraic, geometric and combinatorial\nproperties of a given linear map X. Of basic significance in this theory is the\nfact that the algebraic structures are derived from the geometry (via a\nnon-linear procedure known as \"the least map\"), and that the statistics of the\nalgebraic structures (e.g., the Hilbert series of various polynomial ideals)\nare combinatorial, i.e., computable using a simple discrete algorithm known as\n\"the valuation function\". On the other hand, the theory is somewhat rigid since\nit deals, for the given X, with exactly two pairs each of which is made of a\nnested sequence of three ideals: an external ideal (the smallest), a central\nideal (the middle), and an internal ideal (the largest).\n In this paper we show that the fundamental principles of zonotopal algebra as\ndescribed in the previous paragraph extend far beyond the setup of external,\ncentral and internal ideals by building a whole hierarchy of new\ncombinatorially defined zonotopal spaces.\n", " This report provides a brief review of recently developed extended framework\nfor fundamental physics, designated as Quantum Field Mechanics and including\ncausally complete and intrinsically unified theory of explicitly emerging\nelementary particles, their inherent properties, quantum and relativistic\nbehaviour, interactions and their results. Essential progress with respect to\nusual theory is attained due to the unreduced, nonperturbative analysis of\narbitrary interaction process revealing a qualitatively new phenomenon of\ndynamic multivaluedness of interaction results and their dynamically entangled\nstructure giving rise to universally defined dynamic complexity\n(physics/9806002) and absent in usual, always perturbative and dynamically\nsingle-valued models with the zero value of unreduced complexity (including\nsubstitutes used for \"complexity\", \"chaoticity\", \"self-organisation\", etc.). It\nis shown how the observed world structure, starting from elementary particles\nand their interactions, dynamically emerges from the unreduced interaction\nbetween two initially homogeneous, physically real protofields\n(quant-ph/9902015, quant-ph/9902016). In that way one avoids arbitrary\nimposition of abstract entities and unprovable postulates or \"principles\", let\nalone \"mysteries\", of conventional theory, which now obtain unified, consistent\nand realistic explanation in terms of unreduced dynamic complexity. We complete\nthe description of fundamental world structure emergence and properties by an\noutline of experimental implications and resulting causally substantiated\nchange of research strategy.\n", " We develop an analytical method to prove congruences of the type $$\n\\sum_{k=0}^{(p^r-1)/d}A_kz^k \\equiv\n\\omega(z)\\sum_{k=0}^{(p^{r-1}-1)/d}A_kz^{pk} \\pmod{p^{mr}\\mathbb Z_p[[z]]}\n\\quad \\text{for}\\; r=1,2,\\dots, $$ for primes $p>2$ and fixed integers\n$m,d\\ge1$, where $f(z)=\\sum_{k=0}^\\infty A_kz^k$ is an \"arithmetic\"\nhypergeometric series. Such congruences for $m=d=1$ were introduced by Dwork in\n1969 as a tool for $p$-adic analytical continuation of $f(z)$. Our proofs of\nseveral Dwork-type congruences corresponding to $m\\ge2$ (in other words,\nsupercongruences) are based on constructing and proving their suitable\n$q$-analogues, which in turn have their own right for existence and potential\nfor a $q$-deformation of modular forms and of cohomology groups of algebraic\nvarieties. Our method follows the principles of creative microscoping\nintroduced by us to tackle $r=1$ instances of such congruences; it is the first\nmethod capable of establishing the supercongruences of this type for general\n$r$.\n", " We report some computational results comparing parallel and sequential codes\nfor vertex/facet enumeration problems for convex polyhedra. The problems chosen\nspan the range from simple to highly degenerate polytopes. We tested one code\n(lrs) based on pivoting and four codes (cddr+, ppl, normaliz, PORTA) based on\nthe double description method. normaliz employs parallelization as do the codes\nplrs and mplrs which are based on lrs. We tested these codes using various\nhardware configurations with up to 1200 cores. Major speedups were obtained by\nparallelization, particularly by the code mplrs which uses MPI and can operate\non clusters of machines.\n", " Let $n$ be a positive integer and $I$ a $k$-subset of integers in $[0,n-1]$.\nGiven a $k$-tuple $A=(\\alpha_0, \\cdots, \\alpha_{k-1})\\in \\mathbb{F}^k_{q^n}$,\nlet $M_{A,I}$ denote the matrix $(\\alpha_i^{q^j})$ with $0\\leq i\\leq k-1$ and\n$j\\in I$. When $I=\\{0,1,\\cdots, k-1\\}$, $M_{A,I}$ is called a Moore matrix\nwhich was introduced by E. H. Moore in 1896. It is well known that the\ndeterminant of a Moore matrix equals $0$ if and only if $\\alpha_0,\\cdots,\n\\alpha_{k-1}$ are $\\mathbb{F}_q$-linearly dependent. We call $I$ that satisfies\nthis property a Moore exponent set. In fact, Moore exponent sets are equivalent\nto maximum rank-distance (MRD) code with maximum left and right idealisers over\nfinite fields. It is already known that $I=\\{0,\\cdots, k-1\\}$ is not the unique\nMoore exponent set, for instance, (generalized) Delsarte-Gabidulin codes and\nthe MRD codes recently discovered by Csajb\\'ok, Marino, Polverino and the\nsecond author both give rise to new Moore exponent sets. By using algebraic\ngeometry approach, we obtain an asymptotic classification result: for $q>5$, if\n$I$ is not an arithmetic progression, then there exist an integer $N$ depending\non $I$ such that $I$ is not a Moore exponent set provided that $n>N$.\n", " We study questions in incidence geometry where the precise position of points\nis `blurry' (e.g. due to noise, inaccuracy or error). Thus lines are replaced\nby narrow tubes, and more generally affine subspaces are replaced by their\nsmall neighborhood. We show that the presence of a sufficiently large number of\napproximately collinear triples in a set of points in d dimensional complex\nspace implies that the points are close to a low dimensional affine subspace.\nThis can be viewed as a stable variant of the Sylvester-Gallai theorem and its\nextensions.\n Building on the recently found connection between Sylvester-Gallai type\ntheorems and complex Locally Correctable Codes (LCCs), we define the new notion\nof stable LCCs, in which the (local) correction procedure can also handle small\nperturbations in the euclidean metric. We prove that such stable codes with\nconstant query complexity do not exist. No impossibility results were known in\nany such local setting for more than 2 queries.\n", " The main result of this paper is a proof that a nearly flat, acutely\ntriangulated convex cap C in R^3 has an edge-unfolding to a non-overlapping\npolygon in the plane. A convex cap is the intersection of the surface of a\nconvex polyhedron and a halfspace. \"Nearly flat\" means that every outer face\nnormal forms a sufficiently small angle phi < Phi with the z-axis orthogonal to\nthe halfspace bounding plane. The size of Phi depends on the acuteness gap\nalpha: if every triangle angle is at most pi/2-alpha, then Phi ~= 0.36\nsqrt(alpha) suffices; e.g., for alpha ~= 3deg, Phi = 5deg. Even if C is closed\nto a polyhedron by adding the convex polygonal base under C, this polyhedron\ncan be edge-unfolded without overlap. The proof employs the recent concepts of\nangle-monotone and radially monotone curves. The proof is constructive, leading\nto a polynomial-time algorithm for finding the edge-cuts, at worst O(n^2); a\nversion has been implemented.\n", " We show that the only monic orthogonal polynomials $\\{P_n\\}_{n=0}^{\\infty}$\nthat satisfy\n$$\\pi(x)\\mathcal{D}_{q}^2P_{n}(x)=\\sum_{j=-2}^{2}a_{n,n+j}P_{n+j}(x),\\;\nx=\\cos\\theta,\\;~ a_{n,n-2}\\neq 0,~ n=2,3,\\dots,$$ where $\\pi(x)$ is a\npolynomial of degree at most $4$ and $\\mathcal{D}_{q}$ is the Askey-Wilson\noperator, are Askey-Wilson polynomials and their special or limiting cases.\nThis completes and proves a conjecture by Ismail concerning a structure\nrelation satisfied by Askey-Wilson polynomials. We use the structure relation\nto derive upper bounds for the smallest zero and lower bounds for the largest\nzero of Askey-Wilson polynomials and their special cases.\n", " The Fortran code NMHDECAY computes the masses, couplings and decay widths of\nall Higgs bosons of the NMSSM in terms of its parameters at the electroweak\n(SUSY breaking) scale: the Yukawa couplings lambda and kappa, the soft\ntrilinear terms A_lambda and A_kappa, and tan(beta) and mu_eff = lambda*<S>.\nThe computation of the spectrum includes leading two loop terms, electroweak\ncorrections and propagator corrections. The computation of the decay widths is\ncarried out as in HDECAY, but (for the moment) without three body decays. Each\npoint in parameter space is checked against negative Higgs bosons searches at\nLEP, including unconventional channels relevant for the NMSSM. One version of\nthe program uses generalized SLHA conventions for input and output.\n" ]

[ " Analog beamforming is an attractive and cost-effective solution to exploit\nthe benefits of massive multiple-input-multiple-output systems, by requiring\nonly one up/down-conversion chain. However, the presence of only one chain\nimposes a significant overhead in estimating the channel state information\nrequired for beamforming, when conventional digital channel estimation (CE)\napproaches are used. As an alternative, this paper proposes a novel CE\ntechnique, called periodic analog CE (PACE), that can be performed by analog\nhardware. By avoiding digital processing, the estimation overhead is\nsignificantly lowered and does not scale with number of antennas. PACE involves\nperiodic transmission of a sinusoidal reference signal by the transmitter,\nestimation of its amplitude and phase at each receive antenna via analog\nhardware, and using these estimates for beamforming. To enable such non-trivial\noperation, two reference tone recovery techniques and a novel receiver\narchitecture for PACE are proposed and analyzed, both theoretically and via\nsimulations. Results suggest that in sparse, wide-band channels and above a\ncertain signal-to-noise ratio, PACE aided beamforming suffers only a small loss\nin beamforming gain and enjoys a much lower CE overhead, in comparison to\nconventional approaches. Benefits of using PACE aided beamforming during the\ninitial access phase are also discussed.\n", " The unitary coupled cluster (UCC) approximation is one of the more promising\nwave-function ans\\\"atze for electronic structure calculations on quantum\ncomputers via the variational quantum eigensolver algorithm. However, for large\nsystems with many orbitals, the required number of UCC factors still leads to\nvery deep quantum circuits, which can be challenging to implement. Based on the\nobservation that most UCC amplitudes are small for weakly correlated molecules,\nwe devise an algorithm that employs a Taylor expansion in the small amplitudes,\ntrading off circuit depth for extra measurements. Strong correlations can be\ntaken into account by performing the expansion about a small set of UCC\nfactors, which are treated exactly. Near equilibrium, the Taylor series\nexpansion often works well without the need to include any exact factors; as\nthe molecule is stretched and correlations increase, we find only a small\nnumber of factors need to be treated exactly.\n", " Remarkable effectiveness of the channel or spatial attention mechanisms for\nproducing more discernible feature representation are illustrated in various\ncomputer vision tasks. However, modeling the cross-channel relationships with\nchannel dimensionality reduction may bring side effect in extracting deep\nvisual representations. In this paper, a novel efficient multi-scale attention\n(EMA) module is proposed. Focusing on retaining the information on per channel\nand decreasing the computational overhead, we reshape the partly channels into\nthe batch dimensions and group the channel dimensions into multiple\nsub-features which make the spatial semantic features well-distributed inside\neach feature group. Specifically, apart from encoding the global information to\nre-calibrate the channel-wise weight in each parallel branch, the output\nfeatures of the two parallel branches are further aggregated by a\ncross-dimension interaction for capturing pixel-level pairwise relationship. We\nconduct extensive ablation studies and experiments on image classification and\nobject detection tasks with popular benchmarks (e.g., CIFAR-100, ImageNet-1k,\nMS COCO and VisDrone2019) for evaluating its performance.\n", " We use elliptic Taylor series expansions and interpolation to deduce a number\nof summations for elliptic hypergeometric series. We extend to the well-poised\nelliptic case results that in the $q$-case have previously been obtained by\nCooper and by Ismail and Stanton. We also provide identities involving S.\nBhargava's cubic theta functions.\n", " Recent interest in integrated sensing and communications has led to the\ndesign of novel signal processing techniques to recover information from an\noverlaid radar-communications signal. Here, we focus on a spectral coexistence\nscenario, wherein the channels and transmit signals of both radar and\ncommunications systems are unknown to the common receiver. In this dual-blind\ndeconvolution (DBD) problem, the receiver admits a multi-carrier wireless\ncommunications signal that is overlaid with the radar signal reflected off\nmultiple targets. The communications and radar channels are represented by\ncontinuous-valued range-times or delays corresponding to multiple transmission\npaths and targets, respectively. Prior works addressed recovery of unknown\nchannels and signals in this ill-posed DBD problem through atomic norm\nminimization but contingent on individual minimum separation conditions for\nradar and communications channels. In this paper, we provide an optimal joint\nseparation condition using extremal functions from the Beurling-Selberg\ninterpolation theory. Thereafter, we formulate DBD as a low-rank modified\nHankel matrix retrieval and solve it via nuclear norm minimization. We estimate\nthe unknown target and communications parameters from the recovered low-rank\nmatrix using multiple signal classification (MUSIC) method. We show that the\njoint separation condition also guarantees that the underlying Vandermonde\nmatrix for MUSIC is well-conditioned. Numerical experiments validate our\ntheoretical findings.\n", " In wireless networks, energy consumed for communication includes both the\ntransmission and the processing energy. In this paper, point-to-point\ncommunication over a fading channel with an energy harvesting transmitter is\nstudied considering jointly the energy costs of transmission and processing.\nUnder the assumption of known energy arrival and fading profiles, optimal\ntransmission policy for throughput maximization is investigated. Assuming that\nthe transmitter has sufficient amount of data in its buffer at the beginning of\nthe transmission period, the average throughput by a given deadline is\nmaximized. Furthermore, a \"directional glue pouring algorithm\" that computes\nthe optimal transmission policy is described.\n", " Movable antenna (MA) is a promising technology to improve wireless\ncommunication performance by varying the antenna position in a given finite\narea at the transceivers to create more favorable channel conditions. In this\npaper, we investigate the MA-enhanced multiple-access channel (MAC) for the\nuplink transmission from multiple users each equipped with a single MA to a\nbase station (BS) with a fixed-position antenna (FPA) array. A field-response\nbased channel model is used to characterize the multi-path channel between the\nantenna array of the BS and each user's MA with a flexible position. To\nevaluate the MAC performance gain provided by MAs, we formulate an optimization\nproblem for minimizing the total transmit power of users, subject to a\nminimum-achievable-rate requirement for each user, where the positions of MAs\nand the transmit powers of users, as well as the receive combining matrix of\nthe BS are jointly optimized. To solve this non-convex optimization problem\ninvolving intricately coupled variables, we develop two algorithms based on\nzero-forcing (ZF) and minimum mean square error (MMSE) combining methods,\nrespectively. Specifically, for each algorithm, the combining matrix of the BS\nand the total transmit power of users are expressed as a function of the MAs'\nposition vectors, which are then optimized by using the proposed\nmulti-directional descent (MDD) framework. It is shown that the proposed\nZF-based and MMSE-based MDD algorithms can converge to high-quality suboptimal\nsolutions with low computational complexities. Simulation results demonstrate\nthat the proposed solutions for MA-enhanced multiple access systems can\nsignificantly decrease the total transmit power of users as compared to\nconventional FPA systems employing antenna selection under both perfect and\nimperfect field-response information.\n", " When the users in a MIMO broadcast channel experience different spatial\ntransmit correlation matrices, a class of gains is produced that is denoted\ntransmit correlation diversity. This idea was conceived for channels in which\ntransmit correlation matrices have mutually exclusive eigenspaces, allowing\nnon-interfering training and transmission. This paper broadens the scope of\ntransmit correlation diversity to the case of partially and fully overlapping\neigenspaces and introduces techniques to harvest these generalized gains. For\nthe two-user MIMO broadcast channel, we derive achievable degrees of freedom\n(DoF) and achievable rate regions with/without channel state information at the\nreceiver (CSIR). When CSIR is available, the proposed achievable DoF region is\ntight in some configurations of the number of receive antennas and the channel\ncorrelation ranks. We then extend the DoF results to the $K$-user case by\nanalyzing the interference graph that characterizes the overlapping structure\nof the eigenspaces. Our achievability results employ a combination of product\nsuperposition in the common part of the eigenspaces, and pre-beamforming (rate\nsplitting) to create multiple data streams in non-overlapping parts of the\neigenspaces. Massive MIMO is a natural example in which spatially correlated\nlink gains are likely to occur. We study the achievable downlink sum rate for a\nfrequency-division duplex massive MIMO system under transmit correlation\ndiversity.\n", " For this year's 8TeV run of the LHC we lay out different strategies to search\nfor scalar top pairs. We show results for the hadronic and for the\nsemi-leptonic channels based on hadronic top tagging. For the di-lepton channel\nwe illustrate the impact of transverse mass variables. Each of our\nsignal-to-background ratios ranges around unity for a stop mass around 400GeV.\nThe combined signal significances show that dedicated stop searches are\nbecoming sensitive over a non-negligible part of parameter space.\n", " We give examples showing that the usual Artin Approximation theorems valid\nfor convergent series over a field are no longer true for convergent series\nover a commutative Banach algebra. In particular we construct an example of a\ncommutative integral Banach algebra $R$ such that the ring of formal power\nseries over $R$ is not flat over the ring of convergent power series over $R$.\n", " In this work, we present a phenomenological study of the complete analytical\nsolution to the bound eigenstates and eigenvalues of the Yukawa potential\nobtained previously using the hidden supersymmetry of the system and a\nsystematic expansion of the Yukawa potential in terms of $\\delta=a_{0}/D$,\nwhere $a_{0}$ is the Bohr radius and $D$ is the screening length. The\neigenvalues, $\\epsilon_{nl}(\\delta)$, are given in the form of Taylor series in\n$\\delta$ which can be systematically calculated to the desired order\n$\\delta^{k}$. Coulomb $l$-degeneracy is broken by the screening effects and,\nfor a given $n$, $\\epsilon_{nl}(\\delta)$ is larger for higher values of $l$\nwhich causes the crossing of levels for $n\\ge4$. The convergence radius of the\nTaylor series can be enlarged up to the critical values using the Pad\\'e\napproximants technique which allows us to calculate the eigenvalues with high\nprecision in the whole rage of values of $\\delta$ where bound states exist, and\nto reach a precise determination of the critical screening lengths,\n$\\delta_{nl}$. Eigenstates have a form similar to the solutions of the Coulomb\npotential, with the associated Laguerre polynomials replaced by new polynomials\nof order $\\delta^{k}$ with $r$-dependent coefficients which, in turn, are\npolynomials in $r$. In general we find sizable deviations from the Coulomb\nradial probabilities only for screening lengths close to their critical values.\nWe use these solutions to find the squared absolute value at the origin of the\nwave function for $l=0$, and their derivatives for $l=1$, for the lowest\nstates, as functions of $\\delta$, which enter the phenomenology of dark matter\nbound states in dark gauge theories with a light dark mediator.\n", " For anisotropic superconductors, the gap function belongs to an irreducible\nrepresentation of the point group of crystals, which determines the nodal\nstructure of the gap function. Impurity effects on anisotropic superconductors\nhave been treated by the Born or T-matrix approximation by which only a partial\nsum of the pertutbation series is calculated. Instead, we take into acount all\nterms of the perturbation series. As a result, we find that the introduction of\nnonmagnetic impurities does not change the nodal structure in all cases. We\nalso find how to obtain nodes of the gap function and show the list for\n$O_{h}$, $D_{4h}$, and $D_{6h}$.\n", " The coverage problem in wireless sensor networks deals with the problem of\ncovering a region or parts of it with sensors. In this paper, we address the\nproblem of covering a set of line segments in sensor networks. A line segment `\nis said to be covered if it intersects the sensing regions of at least one\nsensor distributed in that region. We show that the problem of finding the\nminimum number of sensors needed to cover each member in a given set of line\nsegments in a rectangular area is NP-hard. Next, we propose a constant factor\napproximation algorithm for the problem of covering a set of axis-parallel line\nsegments. We also show that a PTAS exists for this problem.\n", " Mixing (or quasirandom) properties of the natural transition matrix\nassociated to a graph can be quantified by its distance to the complete graph.\nDifferent mixing properties correspond to different norms to measure this\ndistance. For dense graphs, two such properties known as spectral expansion and\nuniformity were shown to be equivalent in seminal 1989 work of Chung, Graham\nand Wilson. Recently, Conlon and Zhao extended this equivalence to the case of\nsparse vertex transitive graphs using the famous Grothendieck inequality. Here\nwe generalize these results to the non-commutative, or `quantum', case, where a\ntransition matrix becomes a quantum channel. In particular, we show that for\nirreducibly covariant quantum channels, expansion is equivalent to a natural\nanalog of uniformity for graphs, generalizing the result of Conlon and Zhao.\nMoreover, we show that in these results, the non-commutative and commutative\n(resp.) Grothendieck inequalities yield the best-possible constants.\n", " In this paper, we introduce the notion of Plausible Deniability in an\ninformation theoretic framework. We consider a scenario where an entity that\neavesdrops through a broadcast channel summons one of the parties in a\ncommunication protocol to reveal their message (or signal vector). It is\ndesirable that the summoned party have enough freedom to produce a fake output\nthat is likely plausible given the eavesdropper's observation. We examine three\nvariants of this problem -- Message Deniability, Transmitter Deniability, and\nReceiver Deniability. In the first setting, the message sender is summoned to\nproduce the sent message. Similarly, in the second and third settings, the\ntransmitter and the receiver are required to produce the transmitted codeword,\nand the received vector respectively. For each of these settings, we examine\nthe maximum communication rate that allows a given minimum rate of plausible\nfake outputs. For the Message and Transmitter Deniability problems, we fully\ncharacterise the capacity region for general broadcast channels, while for the\nReceiver Deniability problem, we give an achievable rate region for physically\ndegraded broadcast channels.\n", " Quantitatively predictive models of biomolecular circuits are important tools\nfor the design of synthetic biology and molecular communication circuits. The\ninformation content of typical time-lapse single-cell data for the inference of\nkinetic parameters is not only limited by measurement uncertainty and intrinsic\nstochasticity, but also by the employed perturbations. Novel microfluidic\ndevices enable the synthesis of temporal chemical concentration profiles. The\ninformativeness of a perturbation can be quantified based on mutual\ninformation. We propose an approximate method to perform optimal experimental\ndesign of such perturbation profiles. To estimate the mutual information we\nperform a multivariate log-normal approximation of the joint distribution over\nparameters and observations and scan the design space using Metropolis-Hastings\nsampling. The method is demonstrated by finding optimal perturbation sequences\nfor synthetic case studies on a gene expression model with varying reporter\ncharacteristics.\n", " We prove the neo-classical inequality with the optimal constant, which was\nconjectured by T. J. Lyons [Rev. Mat. Iberoamericana 14 (1998) 215-310]. For\nthe proof, we introduce the fractional order Taylor's series with residual\nterms. Their application to a particular function provides an identity that\ndeduces the optimal neo-classical inequality.\n", " The gains afforded by cloud radio access network (C-RAN) in terms of savings\nin capital and operating expenses, flexibility, interference management and\nnetwork densification rely on the presence of high-capacity low-latency\nfronthaul connectivity between remote radio heads (RRHs) and baseband unit\n(BBU). In light of the non-uniform and limited availability of fiber optics\ncables, the bandwidth constraints on the fronthaul network call, on the one\nhand, for the development of advanced baseband compression strategies and, on\nthe other hand, for a closer investigation of the optimal functional split\nbetween RRHs and BBU. In this chapter, after a brief introduction to signal\nprocessing challenges in C-RAN, this optimal function split is studied at the\nphysical (PHY) layer as it pertains to two key baseband signal processing\nsteps, namely channel estimation in the uplink and channel encoding/ linear\nprecoding in the downlink. Joint optimization of baseband fronthaul compression\nand of baseband signal processing is tackled under different PHY functional\nsplits, whereby uplink channel estimation and downlink channel encoding/ linear\nprecoding are carried out either at the RRHs or at the BBU. The analysis, based\non information-theoretical arguments, and numerical results yields insight into\nthe configurations of network architecture and fronthaul capacities in which\ndifferent functional splits are advantageous. The treatment also emphasizes the\nversatility of deterministic and stochastic successive convex approximation\nstrategies for the optimization of C-RANs.\n", " We improve the Monte-Carlo based QCD sum rules by introducing the rigorous\nH\\\"older-inequality-determined sum rule window and a Breit-Wigner type\nparametrization for the phenomenological spectral function. In this improved\nsum rule analysis methodology, the sum rule analysis window can be determined\nwithout any assumptions on OPE convergence or the QCD continuum.Therefore an\nunbiased prediction can be obtained for the phenomenological parameters (the\nhadronic mass and width etc.). We test the new approach in the $\\rho$ meson\nchannel with re-examination and inclusion of $\\alpha_s$ corrections to\ndimension-4 condensates in the OPE. We obtain results highly consistent with\nexperimental values. We also discuss the possible extension of this method to\nsome other channels.\n", " This paper introduces a simple and computationally efficient algorithm for\nconversion formulae between moments and cumulants. The algorithm provides just\none formula for classical, boolean and free cumulants. This is realized by\nusing a suitable polynomial representation of Abel polynomials. The algorithm\nrelies on the classical umbral calculus, a symbolic language introduced by Rota\nand Taylor in 1994, that is particularly suited to be implemented by using\nsoftware for symbolic computations. Here we give a MAPLE procedure. Comparisons\nwith existing procedures, especially for conversions between moments and free\ncumulants, as well as examples of applications to some well-known distributions\n(classical and free) end the paper.\n", " We study the four channels associated with neutrino-deuteron breakup\nreactions at next-to-next to leading order in effective field theory. We find\nthat the total cross-section is indeed converging for neutrino energies up to\n20 MeV, and thus our calculations can provide constraints on theoretical\nuncertainties for the Sudbury Neutrino Observatory. We stress the importance of\na direct experimental measurement to high precision in at least one channel, in\norder to fix an axial two-body counterterm.\n", " We investigate the secure connectivity of wireless sensor networks under the\npairwise key distribution scheme of Chan et al.. Unlike recent work which was\ncarried out under the assumption of full visibility, here we assume a\n(simplified) communication model where unreliable wireless links are\nrepresented as on/off channels. We present conditions on how to scale the model\nparameters so that the network i) has no secure node which is isolated and ii)\nis securely connected, both with high probability when the number of sensor\nnodes becomes large. The results are given in the form of zero-one laws, and\nexhibit significant differences with corresponding results in the full\nvisibility case. Through simulations these zero-one laws are shown to be valid\nalso under a more realistic communication model, i.e., the disk model.\n", " It is a standing open question to decide whether the Yao-Yao structure for\nunit disk graphs (UDGs) is a length spanner of not. This question is highly\nrelevant to the topology control problem for wireless ad hoc networks. In this\npaper we make progress towards resolving this question by showing that the\nYao-Yao structure is a length spanner for UDGs of bounded aspect ratio. We also\npropose a new local algorithm, called Yao-Sparse-Sink, based on the Yao-Sink\nmethod introduced by Li, Wan, Wang and Frieder, that computes a (1+e)-spanner\nof bounded degree for a given UDG and for given e > 0. The Yao-Sparse-Sink\nmethod enables an efficient local computation of sparse sink trees. Finally, we\nshow that all these structures for UDGs -- Yao, Yao-Yao, Yao-Sink and\nYao-Sparse-Sink -- have arbitrarily large weight.\n", " In this paper, we define an operator function as a series of operators\ncorresponding to the Taylor series representing the function of the complex\nvariable. In previous papers, we considered the case when a function has a\ndecomposition in the Laurent series with the infinite principal part and finite\nregular part. Our central challenge is to improve this result having considered\nas a regular part an entire function satisfying the special condition of the\ngrowth regularity. As an application we consider an opportunity to broaden the\nconditions imposed upon the second term not containing the time variable of the\nevolution equation in the abstract Hilbert space.\n", " A method of proving local continuity of concave functions on convex set\npossessing the $\\mu$-compactness property is presented. This method is based on\na special approximation of these functions.\n The class of $\\mu$-compact sets can be considered as a natural extension of\nthe class of compact metrizable subsets of locally convex spaces, to which\nparticular results well known for compact sets can be generalized.\n Applications of the obtained continuity conditions to analysis of different\nentropic characteristics of quantum systems and channels are considered.\n", " We consider finite-state channels (FSCs) where the channel state is\nstochastically dependent on the previous channel output. We refer to these as\nNoisy Output is the STate (NOST) channels. We derive the feedback capacity of\nNOST channels in two scenarios: with and without causal state information (CSI)\navailable at the encoder. If CSI is unavailable, the feedback capacity is\n$C_{\\text{FB}}= \\max_{P(x|y')} I(X;Y|Y')$, while if it is available at the\nencoder, the feedback capacity is $C_{\\text{FB-CSI}}= \\max_{P(u|y'),x(u,s')}\nI(U;Y|Y')$, where $U$ is an auxiliary RV with finite cardinality. In both\nformulas, the output process is a Markov process with stationary distribution.\nThe derived formulas generalize special known instances from the literature,\nsuch as where the state is i.i.d. and where it is a deterministic function of\nthe output. $C_{\\text{FB}}$ and $C_{\\text{FB-CSI}}$ are also shown to be\ncomputable via convex optimization problem formulations. Finally, we present an\nexample of an interesting NOST channel for which CSI available at the encoder\ndoes not increase the feedback capacity.\n", " We study the impact of the ferroelectric domain wall on the ballistic\nconductance of the single-layer graphene channel in the heterostructure\ngraphene / physical gap / ferroelectric film using Wentzel-Kramers-Brillouin\napproximation. Both self-consistent numerical modeling of the electric field\nand space charge dynamics in the heterostructure and approximate analytical\ntheory show that the domain wall contact with the surface creates p-n junction\nin graphene channel. We calculated that the carriers' concentration induced in\ngraphene by uncompensated ferroelectric dipoles originated from the spontaneous\npolarization abrupt near the surface can reach the values of 1019 m-2 order,\nwhich is in two orders higher than it can be obtained for the gate doped\ngraphene on non-ferroelectric substrates. Therefore we predict that graphene\nchannel with the p-n junction caused by ferroelectric domain wall would be\ncharacterized by rather high ballistic conductivity.\n", " The field equations of Kaluza-Klein (KK) theory have been applied in the\ndomain of cosmology. These equations are solved for a flat universe by taking\nthe gravitational and the cosmological constants as a function of time t. We\nuse Taylor's expansion of cosmological function, $\\Lambda(t)$, up to the first\norder of the time $t$. The cosmological parameters are calculated and some\ncosmological problems are discussed.\n", " The article is based on the differential properties of the inner generators\n(singularities) that occur while handling SP series (sum product series), power\nseries whose Taylor coefficients can be written as sum-product combinations. It\nis an elaboration with numerical details and some explicit calculations of the\nchapter 6 of the joint article with Ecalle on these series. For all polynomial\ninputs $f$ and all monomial inputs $F$, the inner generators corresponding to\nthem verify ordinary differential equations of linear homogeneous type with\npolynomial coefficients. Numerical results strongly suggest the non-existence\nof ODEs for functions $F$ that are rational but not monomial, i.e. not of the\nform ${(1-ax)}^p$, for $p \\in \\mathbb{Z}$.\n", " We present a novel set of methods for analyzing coverage properties in\ndynamic sensor networks. The dynamic sensor network under consideration is\nstudied through a series of snapshots, and is represented by a sequence of\nsimplicial complexes, built from the communication graph at each time point. A\nmethod from computational topology called zigzag persistent homology takes this\nsequence of simplicial complexes as input, and returns a `barcode' containing\nthe birth and death times of homological features in this sequence. We derive\nuseful statistics from this output for analyzing time-varying coverage\nproperties. Further, we propose a method which returns specific representative\ncycles for these homological features, at each point along the birth-death\nintervals. These representative cycles are then used to track coverage holes in\nthe network, and obtain size estimates for individual holes at each time point.\nA weighted barcode, incorporating the size information, is then used as a\nvisual and quantitative descriptor of the dynamic network coverage.\n", " This paper presents a counterexample for the approximation algorithm proposed\nby Durocher and Mehrabi [1] for the general problem of finding a rectangular\npartition of a rectilinear polygon with minimum stabbing number.\n", " We model biochemical signal transduction, based on a ligand-receptor binding\nmechanism, as a discrete-time finite-state Markov channel, which we call the\nBIND channel. We show how to obtain the capacity of this channel, for the case\nof binary output, binary channel state, and arbitrary finite input alphabets.\nWe show that the capacity-achieving input distribution is IID. Further, we show\nthat feedback does not increase the capacity of this channel. We show how the\ncapacity of the discrete-time channel approaches the capacity of Kabanov's\nPoisson channel, in the limit of short time steps and rapid ligand release.\n", " Motivated by recent work of George Andrews and Mircea Merca on the expansion\nof the quotient of the truncation of Euler's pentagonal number series by the\ncomplete series, we provide similar expansion results for averages involving\ntruncations of selected, more general, basic hypergeometric series. In\nparticular, our expansions include new results for averaged truncations of the\nseries appearing in the Jacobi triple product identity, the $q$-Gau{\\ss}\nsummation, and the very-well-poised ${}_5\\phi_5$ summation. We show how special\ncases of our expansions can be used to recover various existing results. In\naddition, we establish new inequalities, such as one for a refinement of the\nnumber of partitions into three different colors.\n", " We consider the problem of assembling a sequence based on a collection of its\nsubstrings observed through a noisy channel. The mathematical basis of the\nproblem is the construction and design of sequences that may be discriminated\nbased on a collection of their substrings observed through a noisy channel. We\nexplain the connection between the sequence reconstruction problem and the\nproblem of DNA synthesis and sequencing, and introduce the notion of a DNA\nstorage channel. We analyze the number of sequence equivalence classes under\nthe channel mapping and propose new asymmetric coding techniques to combat the\neffects of synthesis and sequencing noise. In our analysis, we make use of\nrestricted de Bruijn graphs and Ehrhart theory for rational polytopes.\n", " If a is a point in the domain of convergence of a planar power series f in a\nsingle variable x one con expand f into a planar power series in the variable\n(x-a). One arrives at the notion of planar analytic functions on any domain D\nin the complex plane. It can be described by sections of the sheaf of planar\ngerms. The k-ary exponential series exp(k,x) has infinite radius of\nconvergence. It is possible to define a planar analogue of the classical\nzeta-function. As yet a functional equation for it has not been obtained.\n", " A Chern-Simons current, coming from ghost and anti-ghost fields of\nsupersymmetry theory, can be used to define a spectrum of gene expression in\nnew time series data where a spinor field, as alternative representation of a\ngene, is adopted instead of using the standard alphabet sequence of bases $A,\nT, C, G, U$. After a general discussion on the use of supersymmetry in\nbiological systems, we give examples of the use of supersymmetry for living\norganism, discuss the codon and anti-codon ghost fields and develop an\nalgebraic construction for the trash DNA, the DNA area which does not seem\nactive in biological systems. As a general result, all hidden states of codon\ncan be computed by Chern-Simons 3 forms. Finally, we plot a time series of\ngenetic variations of viral glycoprotein gene and host T-cell receptor gene by\nusing a gene tensor correlation network related to the Chern-Simons current. An\nempirical analysis of genetic shift, in host cell receptor genes with separated\ncluster of gene and genetic drift in viral gene, is obtained by using a tensor\ncorrelation plot over time series data derived as the empirical mode\ndecomposition of Chern-Simons current.\n", " The differential transformation method (DTM) enables the easy construction of\na power-series solution to a nonlinear differential equation. The\nexponentiation operation has not been specifically addressed in the DTM\nliterature, and constructing it iteratively is suboptimal. The recurrence for\nexponentiating a power series by J.C.P. Miller provides a concise\nimplementation of exponentiation by a positive integer for DTM. An\nequally-concise implementation of the exponential function is also provided.\n", " An analysis of known experiments on direct and indirect detection of the\nelectron-type neutrino and antineutrino has been made. The analysis based on a\nnew hypothesis that the observed physical space is formed from a finite set of\nbyuons, \"one-dimensional vectorial objects\". It is shown in the article that\nthe hypothesis for existence of neutrinos advanced by Pauli on the basis of an\nanalysis of the conservation laws, is not unquestionable since the fulfillment\nof these laws may be secured by the physical space itself (physical vacuum)\nbeing the lowest energy state of a discrete oscillating system originating in\nthe course of byuon interaction. This effect is analogous to that of\nM\\\"ossbauer. The direct experiments on detecting neutrinos are explained from\nthe existence of a new information channel due to the uncertainty interval for\ncoordinate of the four-contact byuon interaction forming the interior geometry\nof elementary particles and their properties. Given are also the results of an\nexperiment on observation of cyclic variations of the $\\beta$-decay rate, which\nconfirm the existence of said new information channel.\n", " We consider the system $\\mathcal{F}_4(a,b,c)$ of differential equations\nannihilating Appell's hypergeometric series $F_4(a,b,c;x)$. We find the\nintegral representations for four linearly independent solutions expressed by\nthe hypergeometric series $F_4$. By using the intersection forms of twisted\n(co)homology groups associated with them, we provide the monodromy\nrepresentation of $\\mathcal{F}_4(a,b,c)$ and the twisted period relations for\nthe fundamental systems of solutions of $\\mathcal{F}_4$.\n", " Recently, there has been interest in high-precision approximations of the\nfirst eigenvalue of the Laplace-Beltrami operator on spherical triangles for\ncombinatorial purposes. We compute improved and certified enclosures to these\neigenvalues. This is achieved by applying the method of particular solutions in\nhigh precision, the enclosure being obtained by a combination of interval\narithmetic and Taylor models. The index of the eigenvalue is certified by\nexploiting the monotonicity of the eigenvalue with respect to the domain. The\nclassically troublesome case of singular corners is handled by combining\nexpansions at all corners and an expansion from an interior point. In\nparticular, this allows us to compute 100 digits of the fundamental eigenvalue\nfor the 3D Kreweras model that has been the object of previous efforts.\n", " A new formulation of perturbation theory for a description of the Dirac and\nscalar fields (the Yukawa model) is suggested. As the main approximation the\nself-consistent field model is chosen, which allows in a certain degree to\naccount for the effects caused by the interaction of fields. Such choice of the\nmain approximation leads to a normally ordered form of the interaction\nHamiltonian. Generation of the fermion mass due to the interaction with\nexchange of the scalar boson is investigated. It is demonstrated that, for zero\nbare mass, the fermion can acquire mass only if the coupling constant exceeds\nthe critical value determined by the boson mass. In this connection, the\nproblem of the neutrino mass is discussed.\n", " In our previous study, we showed that the branching process approximation is\nuseful for estimating metabolic robustness, measured using the impact degree.\nBy applying a theory of random family forests, we here extend the branching\nprocess approximation to consider the knockout of {\\it multiple} reactions,\ninspired by the importance of multiple knockouts reported by recent\ncomputational and experimental studies. In addition, we propose a better\ndefinition of the number of offspring of each reaction node, allowing for an\nimproved estimation of the impact degree distribution obtained as a result of a\nsingle knockout. Importantly, our proposed approach is also applicable to\nmultiple knockouts. The comparisons between theoretical predictions and\nnumerical results using real-world metabolic networks demonstrate the validity\nof the modeling based on random family forests for estimating the impact degree\ndistributions resulting from the knockout of multiple reactions.\n", " In the article we develop Euler-Lagrange method and calculate all the roots\nof an arbitrary complex polynomial $P(z)$ on the base of calculation (similar\nto the Bernoulli-Aitken-Nikiporets methods) of the limits of ratios of Hadamard\ndeterminants built by means of coefficients of expansions into Taylor and\nLaurent series of the function~$\\frac{P'(z)}{P(z)}$.\n", " A new modulus of smoothness based on the Euler angles is introduced on the\nunit sphere and is shown to satisfy all the usual characteristic properties of\nmoduli of smoothness, including direct and inverse theorem for the best\napproximation by polynomials and its equivalence to a $K$-functional, defined\nvia partial derivatives in Euler angles. The set of results on the moduli on\nthe sphere serves as a basis for defining new moduli of smoothness and their\ncorresponding $K$-functionals on the unit ball, which are used to characterize\nthe best approximation by polynomials on the ball.\n", " The main result of these notes is an analytical expression for the partition\nfunction of the circular brane model for arbitrary values of the topological\nangle. The model has important applications in condensed matter physics. It is\nrelated to the dissipative rotator (Ambegaokar-Eckern-Schon) model and\ndescribes a ``weakly blocked'' quantum dot with an infinite number of tunneling\nchannels under a finite gate voltage bias. A numerical check of the analytical\nsolution by means of Monte Carlo simulations has been performed recently. To\nderive the main result we study the so-called boundary parafermionic\nsine-Gordon model. The latter is of certain interest to condensed matter\napplications, namely as a toy model for a point junction in the multichannel\nquantum wire.\n", " T. Kaluza has given a criterion for the signs of the power series of a\nfunction that is the reciprocal of another power series. In this note the\nsharpness of this condition is explored and various examples in terms of the\nGaussian hypergeometric series are given. A criterion for the monotonicity of\nthe quotient of two power series due to M. Biernacki and J. Krzy\\.z is applied.\n", " Audio applications involving environmental sound analysis increasingly use\ngeneral-purpose audio representations, also known as embeddings, for transfer\nlearning. Recently, Holistic Evaluation of Audio Representations (HEAR)\nevaluated twenty-nine embedding models on nineteen diverse tasks. However, the\nevaluation's effectiveness depends on the variation already captured within a\ngiven dataset. Therefore, for a given data domain, it is unclear how the\nrepresentations would be affected by the variations caused by myriad\nmicrophones' range and acoustic conditions -- commonly known as channel\neffects. We aim to extend HEAR to evaluate invariance to channel effects in\nthis work. To accomplish this, we imitate channel effects by injecting\nperturbations to the audio signal and measure the shift in the new (perturbed)\nembeddings with three distance measures, making the evaluation domain-dependent\nbut not task-dependent. Combined with the downstream performance, it helps us\nmake a more informed prediction of how robust the embeddings are to the channel\neffects. We evaluate two embeddings -- YAMNet, and OpenL3 on monophonic\n(UrbanSound8K) and polyphonic (SONYC-UST) urban datasets. We show that one\ndistance measure does not suffice in such task-independent evaluation. Although\nFr\\'echet Audio Distance (FAD) correlates with the trend of the performance\ndrop in the downstream task most accurately, we show that we need to study FAD\nin conjunction with the other distances to get a clear understanding of the\noverall effect of the perturbation. In terms of the embedding performance, we\nfind OpenL3 to be more robust than YAMNet, which aligns with the HEAR\nevaluation.\n", " We introduce a mean counting function for Dirichlet series, which plays the\nsame role in the function theory of Hardy spaces of Dirichlet series as the\nNevanlinna counting function does in the classical theory. The existence of the\nmean counting function is related to Jessen and Tornehave's resolution of the\nLagrange mean motion problem. We use the mean counting function to describe all\ncompact composition operators with Dirichlet series symbols on the\nHardy--Hilbert space of Dirichlet series, thus resolving a problem which has\nbeen open since the bounded composition operators were described by Gordon and\nHedenmalm. The main result is that such a composition operator is compact if\nand only if the mean counting function of its symbol satisfies a decay\ncondition at the boundary of a half-plane.\n", " In this paper, we explore quantum interference in molecular conductance from\nthe point of view of graph theory and walks on lattices. By virtue of the\nCayley-Hamilton theorem for characteristic polynomials and the\nCoulson-Rushbrooke pairing theorem for alternant hydrocarbons, it is possible\nto derive a finite series expansion of the Green's function for electron\ntransmission in terms of the odd powers of the vertex adjacency matrix or\nH{\\\"u}ckel matrix. This means that only odd-length walks on a molecular graph\ncontribute to the conductivity through a molecule. Thus, if there are only\neven-length walks between two atoms, quantum interference is expected to occur\nin the electron transport between them. However, even if there are only\nodd-length walks between two atoms, a situation may come about where the\ncontributions to the QI of some odd-length walks are canceled by others,\nleading to another class of quantum interference. For non-alternant\nhydrocarbons, the finite Green's function expansion may include both even and\nodd powers. Nevertheless, QI can in some circumstances come about for\nnon-alternants, from the cancellation of odd and even-length walk terms. We\nreport some progress, but not a complete resolution of the problem of\nunderstanding the coefficients in the expansion of the Green's function in a\npower series of the adjacency matrix, these coefficients being behind the\ncancellations that we have mentioned. And we introduce a perturbation theory\nfor transmission as well as some potentially useful infinite power series\nexpansions of the Green's function.\n", " Goal-oriented dialog has been given attention due to its numerous\napplications in artificial intelligence. Goal-oriented dialogue tasks occur\nwhen a questioner asks an action-oriented question and an answerer responds\nwith the intent of letting the questioner know a correct action to take. To ask\nthe adequate question, deep learning and reinforcement learning have been\nrecently applied. However, these approaches struggle to find a competent\nrecurrent neural questioner, owing to the complexity of learning a series of\nsentences. Motivated by theory of mind, we propose \"Answerer in Questioner's\nMind\" (AQM), a novel information theoretic algorithm for goal-oriented dialog.\nWith AQM, a questioner asks and infers based on an approximated probabilistic\nmodel of the answerer. The questioner figures out the answerer's intention via\nselecting a plausible question by explicitly calculating the information gain\nof the candidate intentions and possible answers to each question. We test our\nframework on two goal-oriented visual dialog tasks: \"MNIST Counting Dialog\" and\n\"GuessWhat?!\". In our experiments, AQM outperforms comparative algorithms by a\nlarge margin.\n" ]

[ " One-clock priced timed games is a class of two-player, zero-sum,\ncontinuous-time games that was defined and thoroughly studied in previous\nworks. We show that one-clock priced timed games can be solved in time m 12^n\nn^(O(1)), where n is the number of states and m is the number of actions. The\nbest previously known time bound for solving one-clock priced timed games was\n2^(O(n^2+m)), due to Rutkowski. For our improvement, we introduce and study a\nnew algorithm for solving one-clock priced timed games, based on the sweep-line\ntechnique from computational geometry and the strategy iteration paradigm from\nthe algorithmic theory of Markov decision processes. As a corollary, we also\nimprove the analysis of previous algorithms due to Bouyer, Cassez, Fleury, and\nLarsen; and Alur, Bernadsky, and Madhusudan.\n", " Reinforcement learning algorithms solve sequential decision-making problems\nin probabilistic environments by optimizing for long-term reward. The desire to\nuse reinforcement learning in safety-critical settings inspires a recent line\nof work on formally constrained reinforcement learning; however, these methods\nplace the implementation of the learning algorithm in their Trusted Computing\nBase. The crucial correctness property of these implementations is a guarantee\nthat the learning algorithm converges to an optimal policy. This paper begins\nthe work of closing this gap by developing a Coq formalization of two canonical\nreinforcement learning algorithms: value and policy iteration for finite state\nMarkov decision processes. The central results are a formalization of Bellman's\noptimality principle and its proof, which uses a contraction property of\nBellman optimality operator to establish that a sequence converges in the\ninfinite horizon limit. The CertRL development exemplifies how the Giry monad\nand mechanized metric coinduction streamline optimality proofs for\nreinforcement learning algorithms. The CertRL library provides a general\nframework for proving properties about Markov decision processes and\nreinforcement learning algorithms, paving the way for further work on\nformalization of reinforcement learning algorithms.\n", " This paper addresses the problem of statistical inference for latent\ncontinuous-time stochastic processes, which is often intractable, particularly\nfor discrete state space processes described by Markov jump processes. To\novercome this issue, we propose a new tractable inference scheme based on an\nentropic matching framework that can be embedded into the well-known\nexpectation propagation algorithm. We demonstrate the effectiveness of our\nmethod by providing closed-form results for a simple family of approximate\ndistributions and apply it to the general class of chemical reaction networks,\nwhich are a crucial tool for modeling in systems biology. Moreover, we derive\nclosed form expressions for point estimation of the underlying parameters using\nan approximate expectation maximization procedure. We evaluate the performance\nof our method on various chemical reaction network instantiations, including a\nstochastic Lotka-Voltera example, and discuss its limitations and potential for\nfuture improvements. Our proposed approach provides a promising direction for\naddressing complex continuous-time Bayesian inference problems.\n", " In hierarchical planning for Markov decision processes (MDPs), temporal\nabstraction allows planning with macro-actions that take place at different\ntime scale in form of sequential composition. In this paper, we propose a novel\napproach to compositional reasoning and hierarchical planning for MDPs under\ntemporal logic constraints. In addition to sequential composition, we introduce\na composition of policies based on generalized logic composition: Given\nsub-policies for sub-tasks and a new task expressed as logic compositions of\nsubtasks, a semi-optimal policy, which is optimal in planning with only\nsub-policies, can be obtained by simply composing sub-polices. Thus, a\nsynthesis algorithm is developed to compute optimal policies efficiently by\nplanning with primitive actions, policies for sub-tasks, and the compositions\nof sub-policies, for maximizing the probability of satisfying temporal logic\nspecifications. We demonstrate the correctness and efficiency of the proposed\nmethod in stochastic planning examples with a single agent and multiple task\nspecifications.\n", " It is known that Dobrushin's ergodicity coefficient is one of the effective\ntools in the investigations of limiting behavior of Markov processes. Several\ninteresting properties of the ergodicity coefficient of a positive mapping\ndefined on ordered Banach space with a base have been studied. In this paper,\nwe consider uniformly mean ergodic and asymptotically stable Markov operators\non ordered Banach spaces. In terms of the ergodicity coefficient, we prove\nuniform mean ergodicity criterion in terms of the ergodicity coefficient.\nMoreover, we develop the perturbation theory for uniformly asymptotically\nstable Markov chains on ordered Banach spaces. In particularly, main results\nopen new perspectives in the perturbation theory for quantum Markov processes\ndefined on von Neumann algebras. Moreover, by varying the Banach spaces one can\nobtain several interesting results in both classical and quantum settings as\nwell.\n", " We provide a rather explicit formula for the resolvent of a~concatenation of\n$N$ processes in terms of their exit laws and certain probability measures\ncharacterizing the way the processes are concatenated. As an application, we\nprove an averaging principle saying that by concatenating asymptotically\nsplittable processes one can approximate Markov chains.\n", " The Turing test for comparing computer performance to that of humans is well\nknown, but, surprisingly, there is no widely used test for comparing how much\nbetter human-computer systems perform relative to humans alone, computers\nalone, or other baselines. Here, we show how to perform such a test using the\nratio of means as a measure of effect size. Then we demonstrate the use of this\ntest in three ways. First, in an analysis of 79 recently published experimental\nresults, we find that, surprisingly, over half of the studies find a decrease\nin performance, the mean and median ratios of performance improvement are both\napproximately 1 (corresponding to no improvement at all), and the maximum ratio\nis 1.36 (a 36% improvement). Second, we experimentally investigate whether a\nhigher performance improvement ratio is obtained when 100 human programmers\ngenerate software using GPT-3, a massive, state-of-the-art AI system. In this\ncase, we find a speed improvement ratio of 1.27 (a 27% improvement). Finally,\nwe find that 50 human non-programmers using GPT-3 can perform the task about as\nwell as--and less expensively than--the human programmers. In this case,\nneither the non-programmers nor the computer would have been able to perform\nthe task alone, so this is an example of a very strong form of human-computer\nsynergy.\n", " Huffman Codes are optimal Instantaneous Fixed-to-Variable (FV) codes in which\nevery source symbol can only be encoded by one codeword. Relaxing these\nconstraints permits constructing better FV codes. More specifically, recent\nwork has shown that AIFV-$m$ codes can beat Huffman coding. AIFV-$m$ codes\nconstruct am $m$-tuple of different coding trees between which the code\nalternates and are only almost instantaneous (AI). This means that decoding a\nword might require a delay of a finite number of bits.\n Current algorithms for constructing optimal AIFV-$m$ codes are iterative\nprocesses that construct progressively \"better sets\" of code trees. The\nprocesses have been proven to finitely converge to the optimal code but with no\nknown bounds on the convergence rate.\n This paper derives a geometric interpretation of the space of binary AIFV-$2$\ncodes, permitting the development of the first polynomially time-bounded\niterative procedures for constructing optimal AIFV codes.\n We first show that a simple binary search procedure can replace the current\niterative process to construct optimal binary AIFV-$2$ codes. We then describe\nhow to frame the problem as a linear programming with an exponential number of\nconstraints but a polynomial-time separability oracle. This permits using the\nGr\\\"otschel, Lov\\'asz and Schrijver ellipsoid method to solve the problem in a\npolynomial number of steps. While more complicated, this second method has the\npotential to lead to a polynomial time algorithm to construct optimal AIFV-$m$\ncodes for general $m$.\n", " This paper proposes, implements, and evaluates a reinforcement learning\n(RL)-based computational framework for automatic mesh generation. Mesh\ngeneration plays a fundamental role in numerical simulations in the area of\ncomputer aided design and engineering (CAD/E). It is identified as one of the\ncritical issues in the NASA CFD Vision 2030 Study. Existing mesh generation\nmethods suffer from high computational complexity, low mesh quality in complex\ngeometries, and speed limitations. These methods and tools, including\ncommercial software packages, are typically semiautomatic and they need inputs\nor help from human experts. By formulating the mesh generation as a Markov\ndecision process (MDP) problem, we are able to use a state-of-the-art\nreinforcement learning (RL) algorithm called \"soft actor-critic\" to\nautomatically learn from trials the policy of actions for mesh generation. The\nimplementation of this RL algorithm for mesh generation allows us to build a\nfully automatic mesh generation system without human intervention and any extra\nclean-up operations, which fills the gap in the existing mesh generation tools.\nIn the experiments to compare with two representative commercial software\npackages, our system demonstrates promising performance with respect to\nscalability, generalizability, and effectiveness.\n", " The aim of this work is to give degree formulas for the generalized Hamming\nweights of evaluation codes and to show lower bounds for these weights. In\nparticular, we give degree formulas for the generalized Hamming weights of\nReed--Muller-type codes, and we determine the minimum distance of toric codes\nover hypersimplices, and the 1st and 2nd generalized Hamming weights of\nsquarefree evaluation codes.\n", " In this article we present a class of codes with few weights arising from\nspecial type of linear sets. We explicitly show the weights of such codes,\ntheir weight enumerator and possible choices for their generator matrices. In\nparticular, our construction yields also to linear codes with three weights\nand, in some cases, to almost MDS codes. The interest for these codes relies on\ntheir applications to authentication codes and secret schemes, and their\nconnections with further objects such as association schemes and graphs.\n", " In this paper, we present an integral Suzuki-type fixed point theorem for\nmultivalued mappings defined on a complete metric space in terms of the\n\\'{C}iri\\'{c} integral contractions. As an application, we will prove an\nexistence and uniqueness theorem for a functional equation arising in dynamic\nprogramming of continuous multistage decision processes.\n", " Convolutional codes are constructed, designed and analysed using row and/or\nblock structures of unit algebraic schemes. Infinite series of such codes and\nof codes with specific properties are derived. Properties are shown\nalgebraically and algebraic decoding methods are derived. For a given rate and\ngiven error-correction capability at each component, convolutional codes with\nthese specifications and with efficient decoding algorithms are constructed.\nExplicit prototype examples are given but in general large lengths and large\nerror capability are achievable. Convolutional codes with efficient decoding\nalgorithms at or near the maximum free distances attainable for the parameters\nare constructible. Unit memory convolutional codes of maximum possible free\ndistance are designed with practical algebraic decoding algorithms.\n LDPC (low density parity check) convolutional codes with efficient decoding\nschemes are constructed and analysed by the methods. Self-dual and\ndual-containing convolutional codes may also be designed by the methods;\ndual-containing codes enables the construction of quantum codes.\n", " A new approach for the approximation of the channel log-likelihood ratio\n(LLR) for wireless channels based on Taylor series is proposed. The\napproximation is applied to the uncorrelated flat Rayleigh fading channel with\nunknown channel state information at the receiver. It is shown that the\nproposed approximation greatly simplifies the calculation of channel LLRs, and\nyet provides results almost identical to those based on the exact calculation\nof channel LLRs. The results are obtained in the context of bit-interleaved\ncoded modulation (BICM) schemes with low-density parity-check (LDPC) codes, and\ninclude threshold calculations and error rate performance of finite-length\ncodes. Compared to the existing approximations, the proposed method is either\nsignificantly less complex, or considerably more accurate.\n", " Solar activity It is vital to accurately predict solar activity, in order to\ndecrease the plausible damage of electronic equipment in the event of a large\nhigh-intensity solar eruption. Recently, we have proposed BELFIS (Brain\nEmotional Learning-based Fuzzy Inference System) as a tool for the forecasting\nof chaotic systems. The structure of BELFIS is designed based on the neural\nstructure of fear conditioning. The function of BELFIS is implemented by\nassigning adaptive networks to the components of the BELFIS structure. This\npaper especially focuses on performance evaluation of BELFIS as a predictor by\nforecasting solar cycles 16 to 24. The performance of BELFIS is compared with\nother computational models used for this purpose, and in particular with\nadaptive neuro-fuzzy inference system (ANFIS).\n", " A new type of solution for the full 3+1 dimensional space-time Schroedinger\nequation is presented here. We consider elegant presentation of the exact\nsolution in a spherical coordinate system, along with the assuming of\nseparation of the two angular co-ordinates from the radial and time variables.\nThe separation of variables follows from an assumed product form of the full\npotential function, which should allow us to reduce the Schroedinger Eq. to\nRiccati ODE in stationary case. If the angular dependence is constant, then the\nazimuthal dependence is linear and the polar dependence is logarithmic. With\nthis reduction the remaining partial differential equation links the radial and\ntime dependence which is, in effect, the standard time-dependent spherical\nradial Schroedinger equation. Besides, we obtain that the time-depended\n2-angles solutions exist if the axis of preferential direction of wave\npropagation is similar to the time arrow in mechanical processes. The other\npossibility is the appropriate phase-shifting of the time-depended solutions. A\nparaxial type approximation is obtained for the solution in the limit in which\nthe polar angle approaches zero.\n", " A spectral Favard theorem for bounded banded lower Hessenberg matrices that\nadmit a positive bidiagonal factorization is found. The large knowledge on the\nspectral and factorization properties of oscillatory matrices leads to this\nspectral Favard theorem in terms of sequences of multiple orthogonal\npolynomials of types I and II with respect to a set of positive\nLebesgue-Stieltjes~measures. Also a multiple Gauss quadrature is proven and\ncorresponding degrees of precision are found.\n This spectral Favard theorem is applied to Markov chains with\n$(p+2)$-diagonal transition matrices, i.e. beyond birth and death, that admit a\npositive stochastic bidiagonal factorization. In the finite case, the\nKarlin-McGregor spectral representation is given. It is shown that the Markov\nchains are recurrent and explicit expressions in terms of the orthogonal\npolynomials for the stationary distributions are given. Similar results are\nobtained for the countable infinite Markov chain. Now the Markov chain is not\nnecessarily recurrent, and it is characterized in terms of the first measure.\nErgodicity of the Markov chain is discussed in terms of the existence of a mass\nat $1$, which is an eigenvalue corresponding to the right and left\neigenvectors.\n", " Let $\\mathcal{C}$ be a set of $m$ by $n$ matrices over $\\mathbb{F}_q$ such\nthat the rank of $A-B$ is at least $d$ for all distinct $A,B\\in \\mathcal{C}$.\nSuppose that $m\\leqslant n$. If $\\#\\mathcal{C}= q^{n(m-d+1)}$, then\n$\\mathcal{C}$ is a maximum rank distance (MRD for short) code. Until 2016,\nthere were only two known constructions of MRD codes for arbitrary $1<d<m-1$.\nOne was found by Delsarte (1978) and Gabidulin (1985) independently, and it was\nlater generalized by Kshevetskiy and Gabidulin (2005). We often call them\n(generalized) Gabidulin codes. Another family was recently obtained by Sheekey\n(2016), and its elements are called twisted Gabidulin codes. In the same paper,\nSheekey also proposed a generalization of the twisted Gabidulin codes. However\nthe equivalence problem for it is not considered, whence it is not clear\nwhether there exist new MRD codes in this generalization. We call the members\nof this putative larger family generalized twisted Gabidulin codes. In this\npaper, we first compute the Delsarte duals and adjoint codes of them, then we\ncompletely determine the equivalence between different generalized twisted\nGabidulin codes. In particular, it can be proven that, up to equivalence,\ngeneralized Gabidulin codes and twisted Gabidulin codes are both proper subsets\nof this family.\n", " We report some computational results comparing parallel and sequential codes\nfor vertex/facet enumeration problems for convex polyhedra. The problems chosen\nspan the range from simple to highly degenerate polytopes. We tested one code\n(lrs) based on pivoting and four codes (cddr+, ppl, normaliz, PORTA) based on\nthe double description method. normaliz employs parallelization as do the codes\nplrs and mplrs which are based on lrs. We tested these codes using various\nhardware configurations with up to 1200 cores. Major speedups were obtained by\nparallelization, particularly by the code mplrs which uses MPI and can operate\non clusters of machines.\n", " In this paper, we define Abelian and consta-Abelian polyadic codes over rings\ndefined as affine algebras over chain rings. For that aim, we use the classical\nconstruction via splittings and multipliers of the underlying Abelian group. We\nalso derive some results on the structure of the associated polyadic codes and\nthe number of codes under these conditions.\n", " Sampling from Gibbs distribution is a central problem in computer science as\nwell as in statistical physics. In this work we focus on the k-colouring model}\nand the hard-core model with fugacity \\lambda when the underlying graph is an\ninstance of Erdos-Renyi random graph G(n,p), where p=d/n and d is fixed.\n We use the Markov Chain Monte Carlo method for sampling from the\naforementioned distributions. In particular, we consider Glauber (block)\ndynamics. We show a dramatic improvement on the bounds for rapid mixing in\nterms of the number of colours and the fugacity for the corresponding models.\nFor both models the bounds we get are only within small constant factors from\nthe conjectured ones by the statistical physicists.\n We use Path Coupling to show rapid mixing. For k and \\lambda in the range of\nour interest the technical challenge is to cope with the high degree vertices,\ni.e. vertices of degree much larger than the expected degree d. The usual\napproach to this problem is to consider block updates rather than single vertex\nupdates for the Markov chain. Taking appropriately defined blocks the effect of\nhigh degree vertices somehow diminishes. However, devising such a construction\nof blocks is a highly non trivial task.\n We develop for a first time a weighting schema for the paths of the\nunderlying graph. Vertices which belong to \"light\" paths, only, can be placed\nat the boundaries of the blocks. Then the tree-like local structure of G(n,d/n)\nallows the construction of simple structured blocks.\n", " Online learning algorithms have become a ubiquitous tool in the machine\nlearning toolbox and are frequently used in small, resource-constraint\nenvironments. Among the most successful online learning methods are Decision\nTree (DT) ensembles. DT ensembles provide excellent performance while adapting\nto changes in the data, but they are not resource efficient. Incremental tree\nlearners keep adding new nodes to the tree but never remove old ones increasing\nthe memory consumption over time. Gradient-based tree learning, on the other\nhand, requires the computation of gradients over the entire tree which is\ncostly for even moderately sized trees. In this paper, we propose a novel\nmemory-efficient online classification ensemble called shrub ensembles for\nresource-constraint systems. Our algorithm trains small to medium-sized\ndecision trees on small windows and uses stochastic proximal gradient descent\nto learn the ensemble weights of these `shrubs'. We provide a theoretical\nanalysis of our algorithm and include an extensive discussion on the behavior\nof our approach in the online setting. In a series of 2~959 experiments on 12\ndifferent datasets, we compare our method against 8 state-of-the-art methods.\nOur Shrub Ensembles retain an excellent performance even when only little\nmemory is available. We show that SE offers a better accuracy-memory trade-off\nin 7 of 12 cases, while having a statistically significant better performance\nthan most other methods. Our implementation is available under\nhttps://github.com/sbuschjaeger/se-online .\n", " Neural codes serve as a language for neurons in the brain. Convex codes,\nwhich arise from the pattern of intersections of convex sets in Euclidean\nspace, are of particular relevance to neuroscience. Not every code is convex,\nhowever, and the combinatorial properties of a code that determine its\nconvexity are still poorly understood. Here we find that a code that can be\nrealized by a collection of open convex sets may or may not be realizable by\nclosed convex sets, and vice versa, establishing that open convex and closed\nconvex codes are distinct classes. We also prove that max intersection-complete\ncodes (i.e. codes that contain all intersections of maximal codewords) are both\nopen convex and closed convex, and provide an upper bound for their minimal\nembedding dimension. Finally, we show that the addition of non-maximal\ncodewords to an open convex code preserves convexity.\n", " We discuss some basic properties of polar convergence in metric spaces. Polar\nconvergence is closely connected with the notion of Delta-convergence of T.C.\nLim known for several years. Possible existence of a topology which induces\npolar convergence is also investigated. Some applications of polar convergence\nfollow.\n", " Let F:R^n -> R be a feedforward ReLU neural network. It is well-known that\nfor any choice of parameters, F is continuous and piecewise (affine) linear. We\nlay some foundations for a systematic investigation of how the architecture of\nF impacts the geometry and topology of its possible decision regions for binary\nclassification tasks. Following the classical progression for smooth functions\nin differential topology, we first define the notion of a generic, transversal\nReLU neural network and show that almost all ReLU networks are generic and\ntransversal. We then define a partially-oriented linear 1-complex in the domain\nof F and identify properties of this complex that yield an obstruction to the\nexistence of bounded connected components of a decision region. We use this\nobstruction to prove that a decision region of a generic, transversal ReLU\nnetwork F: R^n -> R with a single hidden layer of dimension (n + 1) can have no\nmore than one bounded connected component.\n", " Bayesian neural learning feature a rigorous approach to estimation and\nuncertainty quantification via the posterior distribution of weights that\nrepresent knowledge of the neural network. This not only provides point\nestimates of optimal set of weights but also the ability to quantify\nuncertainty in decision making using the posterior distribution. Markov chain\nMonte Carlo (MCMC) techniques are typically used to obtain sample-based\nestimates of the posterior distribution. However, these techniques face\nchallenges in convergence and scalability, particularly in settings with large\ndatasets and network architectures. This paper address these challenges in two\nways. First, parallel tempering is used used to explore multiple modes of the\nposterior distribution and implemented in multi-core computing architecture.\nSecond, we make within-chain sampling schemes more efficient by using Langevin\ngradient information in forming Metropolis-Hastings proposal distributions. We\ndemonstrate the techniques using time series prediction and pattern\nclassification applications. The results show that the method not only improves\nthe computational time, but provides better prediction or decision making\ncapabilities when compared to related methods.\n", " As the use of artificial intelligent (AI) models becomes more prevalent in\nindustries such as engineering and manufacturing, it is essential that these\nmodels provide transparent reasoning behind their predictions. This paper\nproposes the AI-Reasoner, which extracts the morphological characteristics of\ndefects (DefChars) from images and utilises decision trees to reason with the\nDefChar values. Thereafter, the AI-Reasoner exports visualisations (i.e.\ncharts) and textual explanations to provide insights into outputs made by\nmasked-based defect detection and classification models. It also provides\neffective mitigation strategies to enhance data pre-processing and overall\nmodel performance. The AI-Reasoner was tested on explaining the outputs of an\nIE Mask R-CNN model using a set of 366 images containing defects. The results\ndemonstrated its effectiveness in explaining the IE Mask R-CNN model's\npredictions. Overall, the proposed AI-Reasoner provides a solution for\nimproving the performance of AI models in industrial applications that require\ndefect analysis.\n", " Markov chain approximations of symmetric jump processes are investigated.\nTightness results and a central limit theorem are established. Moreover, given\nthe generator of a symmetric jump process with state space $\\mathbbm{R}^d$ the\napproximating Markov chains are constructed explicitly. As a byproduct we\nobtain a definition of the Sobolev space $H^{\\alpha/2}(\\mathbbm{R}^d)$, $\\alpha\n\\in (0,2)$, that is equivalent to the standard one.\n", " In this paper, we have shown the improvement of English to Bharti Braille\nmachine translation system. We have shown how we can improve a baseline NMT\nmodel by adding some linguistic knowledge to it. This was done for five\nlanguage pairs where English sentences were translated into five Indian\nlanguages and then subsequently to corresponding Bharti Braille. This has been\ndemonstrated by adding a sub-module for translating multi-word expressions. The\napproach shows promising results as across language pairs, we could see\nimprovement in the quality of NMT outputs. The least improvement was observed\nin English-Nepali language pair with 22.08% and the most improvement was\nobserved in the English-Hindi language pair with 23.30%.\n", " This study investigates the relationship between tweet sentiment across\ndiverse categories: news, company opinions, CEO opinions, competitor opinions,\nand stock market behavior in the biotechnology sector, with a focus on\nunderstanding the impact of social media discourse on investor sentiment and\ndecision-making processes. We analyzed historical stock market data for ten of\nthe largest and most influential pharmaceutical companies alongside Twitter\ndata related to COVID-19, vaccines, the companies, and their respective CEOs.\nUsing VADER sentiment analysis, we examined the sentiment scores of tweets and\nassessed their relationships with stock market performance. We employed ARIMA\n(AutoRegressive Integrated Moving Average) and VAR (Vector AutoRegression)\nmodels to forecast stock market performance, incorporating sentiment covariates\nto improve predictions. Our findings revealed a complex interplay between tweet\nsentiment, news, biotech companies, their CEOs, and stock market performance,\nemphasizing the importance of considering diverse factors when modeling and\npredicting stock prices. This study provides valuable insights into the\ninfluence of social media on the financial sector and lays a foundation for\nfuture research aimed at refining stock price prediction models.\n", " We study plane trees as a model for RNA secondary structure, assigning energy\nto each tree based on the Nearest Neighbor Thermodynamic Model, and defining a\ncorresponding Gibbs distribution on the trees. Through a bijection between\nplane trees and 2-Motzkin paths, we design a Markov chain converging to the\nGibbs distribution, and establish fast mixing time results by estimating the\nspectral gap of the chain. The spectral gap estimate is established through a\nseries of decompositions of the chain and also by building on known mixing time\nresults for other chains on Dyck paths. In addition to the mathematical aspects\nof the result, the resulting algorithm can be used as a tool for exploring the\nbranching structure of RNA and its dependence on energy model parameters. The\npseudocode implementing the Markov chain is provided in an appendix.\n", " Geometric disentanglement, the separation of latent codes for intrinsic (i.e.\nidentity) and extrinsic(i.e. pose) geometry, is a prominent task for generative\nmodels of non-Euclidean data such as 3D deformable models. It provides greater\ninterpretability of the latent space, and leads to more control in generation.\nThis work introduces a mesh feature, the conformal factor and normal feature\n(CFAN),for use in mesh convolutional autoencoders. We further propose CFAN-VAE,\na novel architecture that disentangles identity and pose using the CFAN\nfeature. Requiring no label information on the identity or pose during\ntraining, CFAN-VAE achieves geometric disentanglement in an unsupervisedway.\nOur comprehensive experiments, including reconstruction, interpolation,\ngeneration, and identity/pose transfer, demonstrate CFAN-VAE achieves\nstate-of-the-art performance on unsupervised geometric disentanglement. We also\nsuccessfully detect a level of geometric disentanglement in mesh convolutional\nautoencoders that encode xyz-coordinates directly by registering its latent\nspace to that of CFAN-VAE.\n", " In this paper we characterize the algebraic structure of two-dimensional\n$(\\alpha,\\beta )$-constacyclic codes of arbitrary length $s.\\ell$ and of their\nduals. For $\\alpha,\\beta \\in \\{1,-1\\}$, we give necessary and sufficient\nconditions for a two-dimensional $(\\alpha,\\beta )$-constacyclic code to be\nself-dual. We also show that a two-dimensional $(\\alpha,1 )$-constacyclic code\n$\\mathcal{C}$ of length $n=s.\\ell$ can not be self-dual if $\\gcd(s,q)= 1$.\nFinally, we give some examples of self-dual, isodual, MDS and quasi-twisted\ncodes corresponding to two-dimensional $(\\alpha,\\beta )$-constacyclic codes.\n", " This paper investigates the propreties of the persistence diagrams stemming\nfrom almost surely continuous random processes on $[0,t]$. We focus our study\non two variables which together characterize the barcode : the number of points\nof the persistence diagram inside a rectangle $]\\!-\\!\\infty,x]\\times\n[x+\\varepsilon,\\infty[$, $N^{x,x+\\varepsilon}$ and the number of bars of length\n$\\geq \\varepsilon$, $N^\\varepsilon$. For processes with the strong Markov\nproperty, we show both of these variables admit a moment generating function\nand in particular moments of every order. Switching our attention to\nsemimartingales, we show the asymptotic behaviour of $N^\\varepsilon$ and\n$N^{x,x+\\varepsilon}$ as $\\varepsilon \\to 0$ and of $N^\\varepsilon$ as\n$\\varepsilon \\to \\infty$. Finally, we study the repercussions of the classical\nstability theorem of barcodes and illustrate our results with some examples,\nmost notably Brownian motion and empirical functions converging to the Brownian\nbridge.\n", " Based on earlier work by Carlen-Maas and the second- and third-named author,\nwe introduce the notion of intertwining curvature lower bounds for graphs and\nquantum Markov semigroups. This curvature notion is stronger than both\nBakry-\\'Emery and entropic Ricci curvature, while also computationally simpler\nthan the latter. We verify intertwining curvature bounds in a number of\nexamples, including finite weighted graphs and graphs with Laplacians admitting\nnice mapping representations, as well as generalized dephasing semigroups and\nquantum Markov semigroups whose generators are formed by commuting jump\noperators. By improving on the best-known bounds for entropic curvature of\ndepolarizing semigroups, we demonstrate that there can be a gap between the\noptimal intertwining and entropic curvature bound. In the case of qubits, this\nimproved entropic curvature bound implies the modified logarithmic Sobolev\ninequality with optimal constant.\n", " The mainstream approach to the development of ontologies is merging\nontologies encoding different information, where one of the major difficulties\nis that the heterogeneity motivates the ontology merging but also limits\nhigh-quality merging performance. Thus, the entity type (etype) recognition\ntask is proposed to deal with such heterogeneity, aiming to infer the class of\nentities and etypes by exploiting the information encoded in ontologies. In\nthis paper, we introduce a property-based approach that allows recognizing\netypes on the basis of the properties used to define them. From an\nepistemological point of view, it is in fact properties that characterize\nentities and etypes, and this definition is independent of the specific labels\nand hierarchical schemas used to define them. The main contribution consists of\na set of property-based metrics for measuring the contextual similarity between\netypes and entities, and a machine learning-based etype recognition algorithm\nexploiting the proposed similarity metrics. Compared with the state-of-the-art,\nthe experimental results show the validity of the similarity metrics and the\nsuperiority of the proposed etype recognition algorithm.\n", " A computational strategy and a collection of codes are presented for studying\nmultiparticle final states in hard hadronic collisions.\n", " Humans spend over 90% of their time in buildings which account for 40% of\nanthropogenic greenhouse gas (GHG) emissions, making buildings the leading\ncause of climate change. To incentivize more sustainable construction, building\ncodes are used to enforce indoor comfort standards and maximum energy use.\nHowever, they currently only reward energy efficiency measures such as\nequipment or envelope upgrades and disregard the actual spatial configuration\nand usage. Using a new hypergraph model that encodes building floorplan\norganization and facilitates automatic geometry creation, we demonstrate that\nspace efficiency outperforms envelope upgrades in terms of operational carbon\nemissions in 72%, 61% and 33% of surveyed buildings in Zurich, New York, and\nSingapore. Automatically generated floorplans for a case study in Zurich\nfurther increase access to daylight by up to 24%, revealing that auto-generated\nfloorplans have the potential to improve the quality of residential spaces in\nterms of environmental performance and access to daylight.\n", " We report on a dramatic improvement of the performance of the classical\ntime-delayed autosynchronization method (TDAS) to control unstable steady\nstates, by applying a time-varying delay in the TDAS control scheme in a form\nof a deterministic or stochastic delay-modulation in a fixed interval around a\nnominal value $T_0$. The successfulness of this variable delay feedback control\n(VDFC) is illustrated by a numerical control simulation of the Lorenz and\nR\\\"{o}ssler systems using three different types of time-delay modulations: a\nsawtooth wave, a sine wave, and a uniform random distribution. We perform a\ncomparative analysis between the VDFC method and the standard TDAS method for a\nsawtooth-wave modulation by analytically determining the domains of control for\nthe generic case of an unstable fixed point of focus type.\n", " To address growth challenges facing large Data Centers and supercomputing\nclusters a new construction is presented for scalable, high throughput, low\nlatency networks. The resulting networks require 1.5-5 times fewer switches,\n2-6 times fewer cables, have 1.2-2 times lower latency and correspondingly\nlower congestion and packet losses than the best present or proposed networks\nproviding the same number of ports at the same total bisection. These advantage\nratios increase with network size. The key new ingredient is the exact\nequivalence discovered between the problem of maximizing network bisection for\nlarge classes of practically interesting Cayley graphs and the problem of\nmaximizing codeword distance for linear error correcting codes. Resulting\ntranslation recipe converts existent optimal error correcting codes into\noptimal throughput networks.\n", " A new object of the probability theory, two-sided chain of events (symbols),\nis introduced. A theory of multi-steps Markov chains with long-range memory,\nproposed earlier in Phys. Rev. E 68, 06117 (2003), is developed and used to\nestablish the correspondence between these chains and two-sided ones. The\nMarkov chain is proved to be statistically equivalent to the definite two-sided\none and vice versa. The results obtained for the binary chains are generalized\nto the chains taking on the arbitrary number of states.\n", " Computing the complexity of Markov bases is an extremely challenging problem;\nno formula is known in general and there are very few classes of toric ideals\nfor which the Markov complexity has been computed. A monomial curve $C$ in\n$\\mathbb{A}^3$ has Markov complexity $m(C)$ two or three. Two if the monomial\ncurve is complete intersection and three otherwise. Our main result shows that\nthere is no $d\\in \\mathbb{N}$ such that $m(C)\\leq d$ for all monomial curves\n$C$ in $\\mathbb{A}^4$. The same result is true even if we restrict to complete\nintersections. We extend this result to all monomial curves in $\\mathbb{A}^n,\nn\\geq 4$.\n", " In the present paper we introduce and study finite point subsets of a special\nkind, called optimum distributions, in the n-dimensional unit cube. Such\ndistributions are closely related with known (delta,s,n)-nets of low\ndiscrepancy. It turns out that optimum distributions have a rich combinatorial\nstructure. Namely, we show that optimum distributions can be characterized\ncompletely as maximum distance separable codes with respect to a non-Hamming\nmetric. Weight spectra of such codes can be evaluated precisely. We also\nconsider linear codes and distributions and study their general properties\nincluding the duality with respect to a suitable inner product. The\ncorresponding generalized MacWilliams identities for weight enumerators are\nbriefly discussed. Broad classes of linear maximum distance separable codes and\nlinear optimum distributions are explicitely constructed in the paper by the\nHermite interpolations over finite fields.\n", " The switch chain is a well-known Markov chain for sampling directed graphs\nwith a given degree sequence. While not ergodic in general, we show that it is\nergodic for regular degree sequences. We then prove that the switch chain is\nrapidly mixing for regular directed graphs of degree d, where d is any positive\ninteger-valued function of the number of vertices. We bound the mixing time by\nbounding the eigenvalues of the chain. A new result is presented and applied to\nbound the smallest (most negative) eigenvalue. This result is a modification of\na lemma by Diaconis and Stroock, and by using it we avoid working with a lazy\nchain. A multicommodity flow argument is used to bound the second-largest\neigenvalue of the chain. This argument is based on the analysis of a related\nMarkov chain for undirected regular graphs by Cooper, Dyer and Greenhill, but\nwith significant extension required.\n", " Model-agnostic feature attributions can provide local insights in complex ML\nmodels. If the explanation is correct, a domain expert can validate and trust\nthe model's decision. However, if it contradicts the expert's knowledge,\nrelated work only corrects irrelevant features to improve the model. To allow\nfor unlimited interaction, in this paper we provide model-agnostic\nimplementations for two popular explanation methods (Occlusion and Shapley\nvalues) to enforce entirely different attributions in the complex model. For a\nparticular set of samples, we use the corrected feature attributions to\ngenerate extra local data, which is used to retrain the model to have the right\nexplanation for the samples. Through simulated and real data experiments on a\nvariety of models we show how our proposed approach can significantly improve\nthe model's performance only by augmenting its training dataset based on\ncorrected explanations. Adding our interactive explanations to active learning\nsettings increases the sample efficiency significantly and outperforms existing\nexplanatory interactive strategies. Additionally we explore how a domain expert\ncan provide feature attributions which are sufficiently correct to improve the\nmodel.\n", " The solution of Dirac's equation for the hydrogen atom according to\nrelativistic wave mechanics yields for each state a vectorial amplitude\nfunction with four components, two large and two small. Each such component has\nits characteristic surface of constant amplitude, of which we plot several\nexamples. For each state of the hydrogen atom there is both a density of\nelectronic charge surrounding the atomic nucleus and an electronic current\ncirculating about the polar axis; the latter generates a magnetic dipolar\nmoment that agrees precisely with experiment.\n", " High-throughput protein interaction detection methods are strongly affected\nby false positive and false negative results. Focused experiments are needed to\ncomplement the large-scale methods by validating previously detected\ninteractions but it is often difficult to decide which proteins to probe as\ninteraction partners. Developing reliable computational methods assisting this\ndecision process is a pressing need in bioinformatics. We show that we can use\nthe conserved properties of the protein network to identify and validate\ninteraction candidates. We apply a number of machine learning algorithms to the\nprotein connectivity information and achieve a surprisingly good overall\nperformance in predicting interacting proteins. Using a 'leave-one-out'\napproach we find average success rates between 20-50% for predicting the\ncorrect interaction partner of a protein. We demonstrate that the success of\nthese methods is based on the presence of conserved interaction motifs within\nthe network. A reference implementation and a table with candidate interacting\npartners for each yeast protein are available at http://www.protsuggest.org\n", " The availability of accurate localization is critical for multi-robot\nexploration strategies; noisy or inconsistent localization causes failure in\nmeeting exploration objectives. We aim to achieve high localization accuracy\nwith contemporary exploration map belief and vice versa without needing global\nlocalization information. This paper proposes a novel simultaneous exploration\nand localization (SEAL) approach, which uses Gaussian Processes (GP)-based\ninformation fusion for maximum exploration while performing communication graph\noptimization for relative localization. Both these cross-dependent objectives\nwere integrated through the Rao-Blackwellization technique. Distributed\nlinearized convex hull optimization is used to select the next-best unexplored\nregion for distributed exploration. SEAL outperformed cutting-edge methods on\nexploration and localization performance in extensive ROS-Gazebo simulations,\nillustrating the practicality of the approach in real-world applications.\n", " Even a single excess electron or ion migrating on the surface of sensitive\nexplosives can catalyze their gradual exothermic decomposition. Mechanisms\nunderlying such a charge-induced gradual thermal decomposition of highly\nsensitive explosives can be different. If sensitive explosive is a polar\nliquid, intense charge-dipole attraction between excess surface charges and\nsurrounding explosive molecules can result in repetitive attempts of solvation\nof these charges by polar explosive molecules. Every attempt of such\nuncompleted nonequilibrium solvation causes local exothermic decomposition of\nthermolabile polar molecules accompanied by further thermal jumping unsolvated\nexcess charges to new surface sites. Thus, ionized mobile hot spots emerge on\ncharged explosive surface. Stochastic migration of ionized hot spots on\nexplosive surface causes gradual exothermic decomposition of the whole mass of\nthe polar explosive. The similar gradual charge-catalyzed exothermic\ndecomposition of both polar and nonpolar highly sensitive explosives can be\nalso caused by intense charge-dipole attacks of surrounding water vapor\nmolecules electrostatically attracted from ambient humid air and strongly\naccelerated towards charged sites on explosive surfaces. Emission of electrons,\nphotons and heat from ionized hot spots randomly migrating on charged surface\nof highly sensitive explosive aerosol nanoparticles converts such particles\ninto the form of short-circuited thermionic nanobatteries.\n", " We present here a set of examples, classes and tools which can be used for\nstatistical analysis in Graphics Processing Units (GPU). This includes binned\nand unbinned maximum likelihood fits, pseudo-experiment generation,\nconvolutions, Markov Chain Monte Carlo method implementations, and limit\nsetting techniques.\n" ]

[ " Time-frequency methods for vibration-based gearbox faults detection have been\nconsidered the most efficient method. Among these methods, continuous wavelet\ntransform (CWT) as one of the best time-frequency method has been used for both\nstationary and transitory signals. Some deficiencies of CWT are problem of\noverlapping and distortion ofsignals. In this condition, a large amount of\nredundant information exists so that it may cause false alarm or\nmisinterpretation of the operator. In this paper a modified method called Exact\nWavelet Analysis is used to minimize the effects of overlapping and distortion\nin case of gearbox faults. To implement exact wavelet analysis, Particle Swarm\nOptimization (PSO) algorithm has been used for this purpose. This method have\nbeen implemented for the acceleration signals from 2D acceleration sensor\nacquired by Advantech PCI-1710 card from a gearbox test setup in Amirkabir\nUniversity of Technology. Gearbox has been considered in both healthy and\nchipped tooth gears conditions. Kernelized Support Vector Machine (SVM) with\nradial basis functions has used the extracted features from exact wavelet\nanalysis for classification. The efficiency of this classifier is then\nevaluated with the other signals acquired from the setup test. The results show\nthat in comparison of CWT, PSO Exact Wavelet Transform has better ability in\nfeature extraction in price of more computational effort. In addition, PSO\nexact wavelet has better speed comparing to Genetic Algorithm (GA) exact\nwavelet in condition of equal population because of factoring mutation and\ncrossover in PSO algorithm. SVM classifier with the extracted features in\ngearbox shows very good results and its ability has been proved.\n", " In this paper, we propose using curvatures in digital space for 3D object\nanalysis and recognition. Since direct adjacency has only six types of digital\nsurface points in local configurations, it is easy to determine and classify\nthe discrete curvatures for every point on the boundary of a 3D object. Unlike\nthe boundary simplicial decomposition (triangulation), the curvature can take\nany real value. It sometimes makes difficulties to find a right value for\nthreshold. This paper focuses on the global properties of categorizing\ncurvatures for small regions. We use both digital Gaussian curvatures and\ndigital mean curvatures to 3D shapes. This paper proposes a multi-scale method\nfor 3D object analysis and a vector method for 3D similarity classification. We\nuse these methods for face recognition and shape classification. We have found\nthat the Gaussian curvatures mainly describe the global features and average\ncharacteristics such as the five regions of a human face. However, mean\ncurvatures can be used to find local features and extreme points such as nose\nin 3D facial data.\n", " Harnessing high-frequency spin dynamics in three-dimensional (3D)\nnanostructures may lead to paradigm-shifting, next generation devices including\nhigh density spintronics and neuromorphic systems. Despite remarkable progress\nin fabrication, the measurement and interpretation of spin dynamics in complex\n3D structures remain exceptionally challenging. Here we take a first step and\nmeasure coherent spin waves within a 3D artificial spin ice (ASI) structure\nusing Brillouin light scattering. The 3D-ASI was fabricated by using a\ncombination of two-photon lithography and thermal evaporation. Two spin-wave\nmodes were observed in the experiment whose frequencies showed a monotonic\nvariation with the applied field strength. Numerical simulations qualitatively\nreproduced the observed modes. The simulated mode profiles revealed the\ncollective nature of the modes extending throughout the complex network of\nnanowires while showing spatial quantization with varying mode quantization\nnumbers. The study shows a well-defined means to explore high-frequency spin\ndynamics in complex 3D spintronic and magnonic structures.\n", " An exact and general expression for the analytic wavelet transform of a\nreal-valued signal is constructed, resolving the time-dependent effects of\nnon-negligible amplitude and frequency modulation. The analytic signal is first\nlocally represented as a modulated oscillation, demodulated by its own\ninstantaneous frequency, and then Taylor-expanded at each point in time. The\nterms in this expansion, called the instantaneous modulation functions, are\ntime-varying functions which quantify, at increasingly higher orders, the local\ndepartures of the signal from a uniform sinusoidal oscillation. Closed-form\nexpressions for these functions are found in terms of Bell polynomials and\nderivatives of the signal's instantaneous frequency and bandwidth. The analytic\nwavelet transform is shown to depend upon the interaction between the signal's\ninstantaneous modulation functions and frequency-domain derivatives of the\nwavelet, inducing a hierarchy of departures of the transform away from a\nperfect representation of the signal. The form of these deviation terms\nsuggests a set of conditions for matching the wavelet properties to suit the\nvariability of the signal, in which case our expressions simplify considerably.\nOne may then quantify the time-varying bias associated with signal estimation\nvia wavelet ridge analysis, and choose wavelets to minimize this bias.\n", " In this paper we extend the atomic decompositions previously obtained for\nBesov spaces related to the forward light cone to general symmetric cones. We\ndo so via wavelet theory adapted to the cone. The wavelet transforms sets up an\nisomorphism between the Besov spaces and certain reproducing kernel function\nspaces on the group, and sampling of the transformed data will provide the\natomic decompositions and frames for the Besov spaces.\n", " This paper is devoted to wavelet analysis on adele ring $\\bA$ and the theory\nof pseudo-differential operators. We develop the technique which gives the\npossibility to generalize finite-dimensional results of wavelet analysis to the\ncase of adeles $\\bA$ by using infinite tensor products of Hilbert spaces. The\nadele ring is roughly speaking a subring of the direct product of all possible\n($p$-adic and Archimedean) completions $\\bQ_p$ of the field of rational numbers\n$\\bQ$ with some conditions at infinity. Using our technique, we prove that\n$L^2(\\bA)=\\otimes_{e,p\\in\\{\\infty,2,3,5,...}}L^2({\\bQ}_{p})$ is the infinite\ntensor product of the spaces $L^2({\\bQ}_{p})$ with a stabilization $e=(e_p)_p$,\nwhere $e_p(x)=\\Omega(|x|_p)\\in L^2({\\bQ}_{p})$, and $\\Omega$ is a\ncharacteristic function of the unit interval $[0,\\,1]$, $\\bQ_p$ is the field of\n$p$-adic numbers, $p=2,3,5,...$; $\\bQ_\\infty=\\bR$. This description allows us\nto construct an infinite family of Haar wavelet bases on $L^2(\\bA)$ which can\nbe obtained by shifts and multi-delations. The adelic multiresolution analysis\n(MRA) in $L^2(\\bA)$ is also constructed. In the framework of this MRA another\ninfinite family of Haar wavelet bases is constructed. We introduce the adelic\nLizorkin spaces of test functions and distributions and give the\ncharacterization of these spaces in terms of wavelet functions. One class of\npseudo-differential operators (including the fractional operator) is studied on\nthe Lizorkin spaces. A criterion for an adelic wavelet function to be an\neigenfunction for a pseudo-differential operator is derived. We prove that any\nwavelet function is an eigenfunction of the fractional operator. These results\nallow one to create the necessary prerequisites for intensive using of adelic\nwavelet bases and pseudo-differential operators in applications.\n", " Comparative analysis of scalar fields is an important problem with various\napplications including feature-directed visualization and feature tracking in\ntime-varying data. Comparing topological structures that are abstract and\nsuccinct representations of the scalar fields lead to faster and meaningful\ncomparison. While there are many distance or similarity measures to compare\ntopological structures in a global context, there are no known measures for\ncomparing topological structures locally. While the global measures have many\napplications, they do not directly lend themselves to fine-grained analysis\nacross multiple scales. We define a local variant of the tree edit distance and\napply it towards local comparative analysis of merge trees with support for\nfiner analysis. We also present experimental results on time-varying scalar\nfields, 3D cryo-electron microscopy data, and other synthetic data sets to show\nthe utility of this approach in applications like symmetry detection and\nfeature tracking.\n", " Let $G$ be a Vilenkin group. In 2008, Y. A. Farkov constructed wavelets on\n$G$ via the multiresolution analysis method. In this article, a\ncharacterization of wavelet sets on $G$ is established, which provides another\nmethod for the construction of wavelets. As an application, the relation\nbetween multiresolution analyses and wavelets determined from wavelet sets is\nalso presented. To some extent, these results positively answer a question\nmentioned by P. Mahapatra and D. Singh in [Bull. Sci. Math. 167 (2021), Paper\nNo. 102945, 20 pp].\n", " We prove a wavelet $T(1)$ theorem for compactness of multilinear\nCalder\\'{o}n-Zygmund (CZ) operators. Our approach characterizes compactness in\nterms of testing conditions and yields a representation theorem for compact CZ\nforms in terms of wavelet and paraproduct forms that reflect the compact nature\nof the operator.\n", " This study showcases the utilisation of OpenCV for extracting features from\nphotos of 2D engineering drawings. These features are then employed to\nreconstruct 3D CAD models in SCAD format and generate 3D point cloud data\nsimilar to LIDAR scans. Many historical mechanical, aerospace, and civil\nengineering designs exist only as drawings, lacking software-generated CAD or\nBIM models. While 2D to 3D conversion itself is not novel, the novelty of this\nwork is in the usage of simple photos rather than scans or electronic\ndocumentation of 2D drawings. The method can also use scanned drawing data.\nWhile the approach is effective for simple shapes, it currently does not\naddress hidden lines in CAD drawings. The Python Jupyter notebook codes\ndeveloped for this purpose are accessible through GitHub.\n", " We present ShapeNet: a richly-annotated, large-scale repository of shapes\nrepresented by 3D CAD models of objects. ShapeNet contains 3D models from a\nmultitude of semantic categories and organizes them under the WordNet taxonomy.\nIt is a collection of datasets providing many semantic annotations for each 3D\nmodel such as consistent rigid alignments, parts and bilateral symmetry planes,\nphysical sizes, keywords, as well as other planned annotations. Annotations are\nmade available through a public web-based interface to enable data\nvisualization of object attributes, promote data-driven geometric analysis, and\nprovide a large-scale quantitative benchmark for research in computer graphics\nand vision. At the time of this technical report, ShapeNet has indexed more\nthan 3,000,000 models, 220,000 models out of which are classified into 3,135\ncategories (WordNet synsets). In this report we describe the ShapeNet effort as\na whole, provide details for all currently available datasets, and summarize\nfuture plans.\n", " In this paper it is shown the performing of an optical transform to state the\nscalar diffraction in the formulation of the wavelet transform and the 'wave\nequations'. From there, a bridge is build between equations of spherical waves\npresented in 1678 by Huygens and the continuous wavelet transform. For such a\npurpose, wavelets are introduced that meet the principles of waves and the\nproperties of wavelets. The following equations are applied in solution to show\na correspondence between the Huygens-Fresnel diffraction and the wavelet\ntransform.\n", " A data structure and toolkit are presented here that allow for the\ndescription and manipulation of mathematical models of three-manifolds and\ntheir interactive display from multiple viewpoints via the OpenGL 3D graphics\npackage. The data structure and vector math package can be extended to support\nan arbitrary number of Euclidean spatial dimensions.\n A model in 4-space is described by its bounding pure simplicial 3-complex. By\nintersecting a 3-flat with this 3-manifold, the algorithm will extract the\nrequested closed pure simplicial 2-complex surface enclosing the desired 3D\nslice. The user can interactively rotate, pan, zoom, and shade arbitrary 3D\nsolid or wire-frame views of the revealed 3D object created by intersection,\nthus exploring both expected and unexpected symmetries or asymmetries in the\nworld of 3-manifolds in 4-space.\n", " Making use of a unified approach to certain classes of induced\nrepresentations, we establish here a number of detailed spectral theoretic\ndecomposition results. They apply to specific problems from non-commutative\nharmonic analysis, ergodic theory, and dynamical systems. Our analysis is in\nthe setting of semidirect products, discrete subgroups, and solenoids. Our\napplications include analysis and ergodic theory of Bratteli diagrams and their\ncompact duals; of wavelet sets, and wavelet representations.\n", " In the light of recent neutrino oscillation and non-oscillation data, we\nrevisit the phenomenological constraints applicable to three observables\nsensitive to absolute neutrino masses: The effective neutrino mass in single\nbeta decay (m_beta); the effective Majorana neutrino mass in neutrinoless\ndouble beta decay (m_2beta); and the sum of neutrino masses in cosmology\n(Sigma). In particular, we include the constraints coming from the first Main\nInjector Neutrino Oscillation Search (MINOS) data and from the Wilkinson\nMicrowave Anisotropy Probe (WMAP) three-year (3y) data, as well as other\nrelevant cosmological data and priors. We find that the largest neutrino\nsquared mass difference is determined with a 15% accuracy (at 2-sigma) after\nadding MINOS to world data. We also find upper bounds on the sum of neutrino\nmasses Sigma ranging from ~2 eV (WMAP-3y data only) to ~0.2 eV (all\ncosmological data) at 2-sigma, in agreement with previous studies. In addition,\nwe discuss the connection of such bounds with those placed on the matter power\nspectrum normalization parameter sigma_8. We show how the partial degeneracy\nbetween Sigma and sigma_8 in WMAP-3y data is broken by adding further\ncosmological data, and how the overall preference of such data for relatively\nhigh values of sigma_8 pushes the upper bound of Sigma in the sub-eV range.\nFinally, for various combination of data sets, we revisit the (in)compatibility\nbetween current Sigma and m_2beta constraints (and claims), and derive\nquantitative predictions for future single and double beta decay experiments.\n", " In the present work we investigate the behavior of all three components of\npersistent spin current in a quasi-periodic Fibonacci ring subjected to Rashba\nand Dresselhaus spin-orbit interactions. Analogous to persistent charge current\nin a conducting ring where electrons gain a Berry phase in presence of magnetic\nflux, spin Berry phase is associated during the motion of electrons in presence\nof a spin-orbit field which is responsible for the generation of spin current.\nThe interplay between two spin-orbit fields along with quasi-periodic Fibonacci\nsequence on persistent spin current is described elaborately, and from our\nanalysis, we can estimate the strength of any one of two spin-orbit couplings\ntogether with on-site energy, provided the other is known.\n", " In the history of cosmology physical paradoxes played important role for\ndevelopment of contemporary world models. Within the modern standard\ncosmological model there are both observational and conceptual cosmological\nparadoxes which stimulate to search their solution. Confrontation of\ntheoretical predictions of the standard cosmological model with the latest\nastrophysical observational data is considered. A review of conceptual problems\nof the Friedmann space expending models, which are in the bases of modern\ncosmological model, is discussed. The main paradoxes, which are discussed in\nmodern literature, are the Newtonian character of the exact Friedmann equation,\nthe violation of the energy conservation within any comoving local volume,\nviolation of the limiting recession velocity of galaxies for the observed high\nredshift objects. Possible observational tests of the nature of the\ncosmological redshift are discussed\n", " We present up to date cosmological bounds on the sum of active neutrino\nmasses as well as on extended cosmological scenarios with additional thermal\nrelics, as thermal axions or sterile neutrino species. Our analyses consider\nall the current available cosmological data in the beginning of year 2014,\nincluding the very recent and most precise Baryon Acoustic Oscillation (BAO)\nmeasurements from the Baryon Oscillation Spectroscopic Survey. In the minimal\nthree active neutrino scenario, we find Sum m_nu < 0.22 eV at 95% CL from the\ncombination of CMB, BAO and Hubble Space Telescope measurements of the Hubble\nconstant. A non zero value for the sum of the three active neutrino masses of\nabout 0.3 eV is significantly favoured at more than 3 standard deviations when\nadding the constraints on sigma_8 and Omega_m from the Planck Cluster catalog\non galaxy number counts. This preference for non zero thermal relic masses\ndisappears almost completely in both the thermal axion and massive sterile\nneutrino schemes. Extra light species contribute to the effective number of\nrelativistic degrees of freedom, parameterised via Neff. We found that when the\nrecent detection of B mode polarization from the BICEP2 experiment is\nconsidered, an analysis of the combined CMB data in the framework of LCDM+r\nmodels gives Neff=4.00pm0.41, suggesting the presence of an extra relativistic\nrelic at more than 95 % c.l. from CMB-only data.\n", " Soft diffraction phenomena in elastic nucleon scattering are considered from\nthe viewpoint of the spin dependence of the interaction potential.\nSpin-dependent pomeron effects are analyzed for elastic $pp$ scattering,\nspin-dependent differential cross sections and spin correlation parameters are\ncalculated. The spin correlation parameter $A_N$ is examined on the basis of\nexperimental data from $\\sqrt{s} = 4.9 \\ GeV$ up to $23.4 \\ $GeV in the\nframework of the extended High Energy Generalized Structure (HEGS) model. It is\nshown that the existing experimental data of proton-proton and\nproton-antiproton elastic scattering at high energy in the region of the\ndiffraction minimum and at large momentum transfer give the support of the\nexistence of the energy-independent part of the hadron spin flip amplitude.\n", " In this work, we consider the propagation of scalar particles in a\nhigher-dimensional Schwarzschild-de-Sitter black-hole spacetime, both on the\nbrane and in the bulk. Our analysis applies for arbitrary partial modes and for\nboth minimal and non-minimal coupling of the scalar field. A general expression\nfor the greybody factor is analytically derived in each case, and its\nlow-energy behaviour is studied in detail. Its profile in terms of scalar\nproperties (angular-momentum number and non-minimal coupling parameter) and\nspacetime properties (number of extra dimensions and cosmological constant) is\nthoroughly investigated. In contrast to previous studies, the effect of the\ncosmological constant is taken into account both close to and far away from the\nblack-hole horizon. The dual role of the cosmological constant, that may act\neither as a helping agent to the emission of scalar particles or as a deterring\neffect depending on the value of the non-minimal coupling parameter, is also\ndemonstrated.\n", " Motivated by wavelet analysis, we prove that there is a one-to-one\ncorrespondence between the following data: Solutions to $R(h)=h$ where $R$ is a\ncertain non-positive Ruelle transfer operator; Operators that intertwine a\ncertain class of representations of the $C^*$-algebra $\\mathfrak{A}_N$ on two\nunitary generators $U$, $V$ subject to the relation $$UVU^{-1}=V^N$$ This\ncorrespondence enables us to give a criterion for the biorthogonality of a pair\nof scaling functions and calculate all solutions of the equation $R(h)=h$ in\nsome concrete cases.\n", " While texture analysis is largely addressed for images, the comparison of the\ngeometric reliefs on surfaces embedded in the 3D space is still an open\nchallenge. Starting from the Local Binary Pattern (LBP) description originally\ndefined for images, we introduce the edge-Local Binary Pattern (edgeLBP) as a\nlocal description able to capture the evolution of repeated, geometric patterns\non surface meshes. Our extension is independent of the surface representation,\nindeed the edgeLBP is able to deal with surface tessellations characterized by\nnon-uniform vertex distributions and different types of faces, such as\ntriangles, quadrangles and, in general, convex polygons. Besides the desirable\nrobustness properties the edgeLBP exhibits over a number of examples, we show\nhow this description performs well for 3D pattern retrieval and compare our\nperformances with the participants to a recent 3D pattern retrieval and\nclassification contest.\n", " The first part of this article is devoted to a brief review of the results\nabout representation theory of the spin group Spin(m) from the point of view of\nClifford analysis. In the second part we are interested in Clifford-valued\nfunctions and wavelets on the sphere. The connection of representations of\nSpin(m) and the concept of diffusive wavelets leads naturally to investigations\nof a modified diffusion equation on the sphere, that makes use of the Gamma\noperator. We will achieve to obtain Clifford-valued diffusion wavelets with\nrespect to a modified diffusion operator.\n Since we are able to characterize all representations of Spin(m) and even to\nobtain all eigenvectors of the (by representation) regarded Casimir operator in\nrepresentation spaces, it seems appropriate to look at functions on Spin(m)\ndirectly. Concerning this, our aim shall be to formulate eigenfunctions for the\nLaplace-Beltrami operator on Spin(m) and give the series expansion of the heat\nkernel on Spin(m) in terms of eigenfunctions.\n", " This paper introduces the concept of Bessel multipliers. These operators are\ndefined by a fixed multiplication pattern, which is inserted between the\nAnalysis and synthesis operators. The proposed concept unifies the approach\nused for Gabor multipliers for arbitrary analysis/synthesis systems, which form\nBessel sequences, like wavelet or irregular Gabor frames. The basic properties\nof this class of operators are investigated. In particular the implications of\nsummability properties of the symbol for the membership of the corresponding\noperators in certain operator classes are specified. As a special case the\nmultipliers for Riesz bases are examined and it is shown that multipliers in\nthis case can be easily composed and inverted. Finally the continuous\ndependence of a Bessel multiplier on the parameters (i.e. the involved\nsequences and the symbol in use) is verified, using a special measure of\nsimilarity of sequences.\n", " We study theoretically the spin transport in a nonmagnetic metal connected to\nferromagnetic injector and detector electrodes. We derive a general expression\nfor the spin accumulation signal which covers from the metallic to the\ntunneling regime. This enables us to discuss recent controversy on spin\ninjection and detection experiments. Extending the result to a superconducting\ndevice, we find that the spin accumulation signal is strongly enhanced by\nopening of the superconducting gap since a gapped superconductor is a low\ncarrier system for spin transport but not for charge. The enhancement is also\nexpected in semiconductor devices.\n", " Here, we investigate spin Hall angle of Pt in Ni80Fe20/Pt bilayer system by\nusing a broadband spin pumping and inverse spin Hall effect measurement. An\nout-of-plane excitation geometry with application of external magnetic field\nperpendicular to the charge current direction is utilized in order to suppress\nunwanted galvanomagnetic effects. Magnetization precession angle on\nferromagnetic resonance for wide excitation frequency range (4-14 GHz) is\nestimated from the rectification voltage of anisotropic magnetoresistance (AMR)\nand a conventional method of using microwave power in a coplanar waveguide. A\nmarked difference in the precession angle profiles for the different methods is\nobserved, resulting in the large variation in estimated values of spin current\ndensity at Ni80Fe20/Pt interface. The frequency dependence of the spin current\ndensity estimated using AMR effect is found to be similar to that of the\ninverse spin Hall voltage. We obtain the frequency-invariant spin Hall angle of\n0.067.\n", " A generalized Frenkel condition is proposed for use in spin hydrodynamics to\nrelate the spin density and spin polarization (or spin chemical potential)\ntensors. It allows for independent treatment of electric- and magnetic-like\ncomponents of the spin density tensor, which helps to fulfill the stability\nconditions recently derived in the literature. The generalized Frenkel\ncondition extrapolates between the original Frenkel condition, where only the\nmagnetic-like part of the spin tensor is present, and the case where the spin\ndensity tensor is directly proportional to the spin polarization tensor. We\nalso demonstrate that our approach is supported by the result of a microscopic\ncalculation.\n", " Spatial relationships between objects provide important information for\ntext-based image retrieval. As users are more likely to describe a scene from a\nreal world perspective, using 3D spatial relationships rather than 2D\nrelationships that assume a particular viewing direction, one of the main\nchallenges is to infer the 3D structure that bridges images with users' text\ndescriptions. However, direct inference of 3D structure from images requires\nlearning from large scale annotated data. Since interactions between objects\ncan be reduced to a limited set of atomic spatial relations in 3D, we study the\npossibility of inferring 3D structure from a text description rather than an\nimage, applying physical relation models to synthesize holistic 3D abstract\nobject layouts satisfying the spatial constraints present in a textual\ndescription. We present a generic framework for retrieving images from a\ntextual description of a scene by matching images with these generated abstract\nobject layouts. Images are ranked by matching object detection outputs\n(bounding boxes) to 2D layout candidates (also represented by bounding boxes)\nwhich are obtained by projecting the 3D scenes with sampled camera directions.\nWe validate our approach using public indoor scene datasets and show that our\nmethod outperforms baselines built upon object occurrence histograms and\nlearned 2D pairwise relations.\n", " The spinors of the group Spin($N$) of rotations in $N$ spacetime dimensions\nare indexed by a bitcode with [$N$/2] bits. A well-known promising grand\nunified group that contains the standard-model group is Spin(10). Fermions in\nthe standard model are described by five bits $yzrgb$, consisting of two weak\nbits $y$ and $z$, and three color bits $r$, $g$, $b$. If a sixth bit $t$ is\nadded, necessary to accommodate a time dimension, then the enlarged Spin(11,1)\ngeometric algebra contains the standard model and Dirac algebras as commuting\nsubalgebras, unifying the four forces of Nature. There is a unique minimal\nsymmetry-breaking chain and associated multiplet of Higgs fields that breaks\nSpin(11,1) to the standard model. Unification to the Pati-Salam group\nSpin(4)$_w {\\times}$ Spin(6)$_c$ is predicted at $10^{12}\\,$GeV, and grand\nunification at $10^{15}\\,$GeV. The grand Higgs field breaks $t$-symmetry, can\ndrive cosmological inflation, and generates a large Majorana mass for the\nright-handed neutrino by flipping its $t$-bit. The electroweak Higgs field\nbreaks $y$-symmetry, and generates masses for fermions by flipping their\n$y$-bit.\n", " Inherent asymmetry in the tunneling barriers of few-electron quantum dots\ninduces intrinsically different tunneling currents for forward and reverse\nsource-drain biases in the non-linear transport regime. Here we show that in\naddition to spin selection rules, overlap matrix elements between many-body\nstates are crucial for the correct description of tunneling transmission\nthrough quantum dots at large magnetic fields. Signatures of excited\n(N-1)-electron states in the transport process through the N-electron system\nare clearly identified in the measured transconductances. Our analysis clearly\nconfirms the validity of single-electron quantum transport theory in quantum\ndots.\n", " In a first step, we explore the discovery and analysis potentials of the HERA\ncollider, with and without polarized beams, in the search for electron-quark\ncompositeness in the neutral current channel. Then we study the parity\nviolating effects, for jet production in polarized $pp$ collisions at RHIC,\nwhich could be due to the presence of quark subconstituents or new massive\ngauge bosons. We emphasize that the measurement of spin asymmetries in such a\npolarized context could give some crucial informations on the chiral structure\nof these hypothetical new interactions.\n", " As a generalization of orthonormal wavelets in $L_2(R)$, tight framelets\n(also called tight wavelet frames) are of importance in wavelet analysis and\napplied sciences due to their many desirable properties in applications such as\nimage processing and numerical algorithms. Tight framelets are often derived\nfrom particular refinable functions satisfying certain stringent conditions.\nConsequently, a large family of refinable functions cannot be used to construct\ntight framelets. This motivates us to introduce the notion of a quasi-tight\nframelet, which is a dual framelet but behaves almost like a tight framelet. It\nturns out that the study of quasi-tight framelets is intrinsically linked to\nthe problem of the generalized matrix spectral factorization for matrices of\nLaurent polynomials. In this paper, we provide a systematic investigation on\nthe generalized matrix spectral factorization problem and compactly supported\nquasi-tight framelets. As an application of our results on generalized matrix\nspectral factorization for matrices of Laurent polynomials, we prove in this\npaper that from any arbitrary compactly supported refinable function in\n$L_2(R)$, we can always construct a compactly supported one-dimensional\nquasi-tight framelet having the minimum number of generators and the highest\npossible order of vanishing moments. Our proofs are constructive and\nsupplemented by step-by-step algorithms. Several examples of quasi-tight\nframelets will be provided to illustrate the theoretical results and algorithms\ndeveloped in this paper.\n", " In this paper, asymptotic behavior of the uncertainty product of\nGauss-Weierstrass wavelet is investigated. It is shown that the uncertainty\nproduct is bounded from above, a feature that distinguishes Gauss-Weierstrass\nwavelet among other wavelet families.\n", " Curvature principal directions on geometric surfaces are a ubiquitous concept\nof Geometry Processing techniques. However they only account for order 2\ndifferential quantities, oblivious of higher order differential behaviors. In\nthis paper, we extend the concept of principal directions to higher orders for\nsurfaces in R^3 by considering symmetric differential tensors. We further show\nhow they can be explicitly approximated on point set surfaces and that they\nconvey valuable geometric information, that can help the analysis of 3D\nsurfaces.\n", " We construct new classes of cosmological solutions in modified and Einstein\ngravity theories encoding space-time quasicrystal, STQC, configurations\nmodelled by nonlinear self-organized and pattern forming quasi-periodic\nstructures. Such solutions are defined by generic off-diagonal locally\nanisotropic and inhomogeneous metrics depending via generating and integration\nfunctions on all spacetime coordinates. There are defined nonholonomic\nvariables and conditions for the generating/integration functions and sources\nfor effective descriptions, or approximations, as \"quasi\"\nFriedmann-Lama\\^itre-Robertson-Walker (FLRW) metrics. Such (off-) diagonal\nSTQC-FLRW configurations contain memory on nonlinear classical and/or quantum\ninteractions and may describe new acceleration cosmology scenarios. For special\ntime-periodic conditions on nonlinear gravitational and matter field\ninteractions, we can model at cosmological scales certain analogous of time\ncrystal like structures originally postulated by Frank Wilczek in condensed\nmatter physics. We speculate how STQC quasi-FLRW configurations could explain\nmodern cosmology data and provide viable descriptions for the inflation and\nstructure formation in our Universe. Finally, it is discussed systematically\nand critically how a unified description of inflation with dark energy era can\nbe explained by (modified) cosmological STQC-scenarios.\n", " In this work interacting Umami Chaplygin gas has been studied in flat FRW\nmodel of universe in context of it's thermodynamic and dynamical behaviour. In\nparticular, considering Umami fluid as dark energy interacting with dark\nmatter, irreversible thermodynamics has been studied both for apparent and\nevent horizon as bounding horizon in two separate cases. Also the model has\nbeen investigated in purview of dynamical systems analysis by converting the\ncosmological evolution equations to an autonomous system of ordinary\ndifferential equations. With some restrictions on model parameter $\\omega$ and\ncoupling parameter $\\lambda$, some cosmologically interesting critical points\ndescribing late time accelerated evolution of the universe attracted by\ncosmological constant and accelerated scaling attractor in quintessence era\nhave been found to alleviate coincidence problem.\n", " We formulate the Bargman-Michel-Telegdi (BMT) equation for electron spin\nmotion in a plane wave and in the Dirac delta-function pulse. We compare the\nBMT solution with the Wolkow solution of the Dirac equation. The Wolkow\nsolution for the spin is not identical with the solution following from the BMT .\n", " In the new framework of gravitational quantum field theory (GQFT) with spin\nand scaling gauge invariance developed in Phys. Rev. D\\textbf{93} (2016)\n024012-1~\\cite{Wu:2015wwa}, we make a perturbative expansion for the full\naction in a background field which accounts for the early inflationary\nuniverse. We decompose the bicovariant vector fields of gravifield and spin\ngauge field with Lorentz and spin symmetries SO(1,3) and SP(1,3) in biframe\nspacetime into SO(3) representations for deriving the propagators of the basic\nquantum fields and extract their interaction terms. The leading order Feynman\nrules are presented. A tree-level 2 to 2 scattering amplitude of the Dirac\nfermions, through a gravifield and a spin gauge field, is calculated and\ncompared to the Born approximation of the potential. It is shown that the\nNewton's gravitational law in the early universe is modified due to the\nbackground field. The spin dependence of the gravitational potential is\ndemonstrated.\n", " Reconstructing and processing the 3D objects are popular activities in the\nresearch field of computer graphics, image processing and computer vision. The\n3D objects are processed based on the methods like geometric modeling, a branch\nof applied mathematics and computational geometry, or the machine learning\nalgorithms based on image processing. The computation of geometrical objects\nincludes processing the curves and surfaces, subdivision, simplification,\nmeshing, holes filling, reconstructing, and refining the 3D surface objects on\nboth point cloud data and triangular mesh. While the machine learning methods\nare developed using deep learning models. With the support of 3D laser scan\ndevices and Lidar techniques, the obtained dataset is close to original shape\nof the real objects. Besides, the photography and its application based on the\nmodern techniques in recent years help us collect data and process the 3D\nmodels more precise. This article proposes an improved method for filling holes\non the 3D object surface based on an automatic segmentation. Instead of filling\nthe hole directly as the existing methods, we now subdivide the hole before\nfilling it. The hole is first determined and segmented automatically based on\ncomputation of its local curvature. It is then filled on each part of the hole\nto match its local curvature shape. The method can work on both 3D point cloud\nsurfaces and triangular mesh surface. Comparing to the state of the art\nmethods, our proposed method obtained higher accuracy of the reconstructed 3D\nobjects.\n", " We consider non-perturbative effects at short distances in theories with\nconfinement. The analysis is straightforward within the Abelian models in which\nthe confinement arises on classical level. In all cases considered (compact\nU(1) in 3D and 4D, dual Abelian Higgs model) there are non-perturbative\ncontributions associated with short distances which are due to topological\ndefects. In QCD case, both classical and quantum effects determine the role of\nthe topological defects and the theoretical analysis has not been completed so\nfar. Generically, the topological defects would result in 1/Q^2 corrections\ngoing beyond the standard Operator Product Expansion. We review existing data\non the power corrections and find that the data favor existence of the novel\ncorrections, at least at the mass scale of (1-2) GeV. We indicate crucial\nexperiments which could further clarify the situation on the phenomenological\nside.\n", " The theoretical prediction of topological insulators in 2007 triggered\ntremendous interest. They are of fundamental interest because of their\ntopological twist in k-space, which comes along with unidirectional,\nspin-polarized surface-state currents, required for spin-optoelectronics. This\nproperty makes topological insulators on one hand perfect materials for\noptically generated, ultrafast spin-bunches spin-current sources for the\ngeneration of THz radiation. On the other hand, those spin-polarized\nsurface-state currents when generated by a voltage lead to large spin Hall\neffects, or when generated by a temperature gradient to the thermal analogue,\nthe spin Nernst effect. Both mutually convert charge/ heat currents into\ntransverse spin currents leading to spin accumulations. By connecting both\nresearch fields, we show the evidence of heat-transport related spin Hall\neffects that can be extracted from opto-transport experiments. This heat-driven\nspin Nernst effect drives a transverse spin-current and affects the optical\nspin-orientation in the three-dimensional topological insulator. This manifests\nas a modification of the circular polarization-dependent photocurrent. We\nilluminate the detailed thermocurrent distribution, including the influence of\nedges and contacts, in spatially resolved current maps.\n", " Generically,the classical evolution of the inflaton has a brief fast roll\nstage that precedes the slow roll regime. The fast roll stage leads to a purely\nattractive potential for the curvature and tensor perturbations (this potential\nis purely repulsive in slowroll). This attractive potential depresses the CMB\nquadrupole moment for the curvature and B-mode spectra. A single new\ncosmological parameter emerges: the comoving wavenumber k_1 characteristic\nscale of this attractive potential.k_1 happens to exit the horizon precisely at\nthe transition from fast to slowroll.The fastroll stage dynamically modifies\nthe initial power spectrum by a transfer function D(k). We compute D(k) by\nsolving the inflaton evolution equations. D(k) suppresses the primordial power\nfor k < k_1 and enjoys the scaling property D(k) = Psi(k/k_1) where Psi(x) is\nan universal function. We perform a MCMC analysis of the WMAP/SDSS data\nincluding the fast-roll stage and find k_1 = 0.266/Gpc. The quadrupole mode k_Q\n= 0.242/Gpc exits the horizon 1/10 of an efold before k_1. We compare the\nfastroll fit with a fit with a sharp lower cutoff on the primordial power.\nFastroll provides a slightly better fit than a sharp cutoff for the TT,TE and\nEE modes.Moreover, our fits provide non-zero lower bounds for r, while for the\nother cosmological parameters we essentially get those of the LCDM model. The\nfact that k_Q exits the horizon before the slowroll stage implies an upper\nbound in the total number of efolds N_{tot} during inflation.Combining this\nwith estimates during the radiation dominated era we obtain N_{tot} ~ 66, with\nthe bounds 62 < N_{tot} < 82.We repeated the analysis with the\nWMAP-5/ACBAR-2007/SDSS data confirming the overall picture.\n", " We update the constraints on the time variation of the fine structure\nconstant $\\alpha$ and the electron mass $m_e$, using the latest CMB data,\nincluding the 7-yr release of WMAP. We made statistical analyses of the\nvariation of each one of the constants and of their joint variation, together\nwith the basic set of cosmological parameters. We used a modified version of\nCAMB and COSMOMC to account for these possible variations. We present bounds on\nthe variation of the constants for different data sets, and show how results\ndepend on them. When using the latest CMB data plus the power spectrum from\nSloan Digital Sky Survey LRG, we find that $\\alpha / \\alpha_0=0.986 \\pm 0.007$\nat 1-$\\sigma$ level, when the 6 basic cosmological parameters were fitted, and\nonly variation in $\\alpha$ was allowed. The constraints in the case of\nvariation in both constants are $ \\alpha / \\alpha_0= 0.986 \\pm 0.009$ and $m_e\n/ m_{e0} = 0.999 \\pm 0.035$. In the case of only variation in $m_e$, the bound\nis $m_e /m_{e0}=0.964 \\pm 0.025$.\n", " A simple differential analysis of issue of the correspondence between notion\nof geodesics in gravitation theory of GTR and straights of inertial motion in\nthe Minkowski space-time discovers that, conventional certification of the\ngeodesics in GTR is not compatible with the existence of the\nRiemann-Christoffel curvature tensor (RCT). We show that, resolution of the\ncontradiction consists of a natural extension of the Christoffel symbols in the\ngravitation dynamic law to the complete connectedness form which includes a\ntriadic asymmetric tensor (named the moderator). The correspondent Riemann\nsupertensor form, unavoidably annihilating by certification of 4-vectors of\nparticle momentum and spin, gives birth to torsion (skew-symmetric part of the\nmoderator) and the gravitensor (the even-symmetric part); both arrive connected\nto the RCT and become an indispensable integral part of structure of the\ngravitational field. The equivalence principle still actual while it becomes\nenriched in the content. The Einstein-Hilbert law of the metric to matter\nconnection remains unchanged at the produced correction of the gravitational\ndynamics. Our analysis results in the gravitensor addition to Christoffels in\nequation for geodesics, modified equations for 4-vector of particle spin with\ncontribution from torsion, and renormalization of metric in the update dynamic\nconcept of the GTR. We pay attention to possible implication of torsion in the\nelementary interactions.\n", " The cosmological concordance model is consistent with all available\nobservational data, including the apparent distance and redshift relationship\nfor distant supernovae, but it is curious how the Milne cosmological model is\nable to make predictions that are similar to this preferred General\nRelativistic model. Milne's cosmological model is based solely on Special\nRelativity and presumes a completely incompatible redshift mechanism; how then\ncan the predictions be even remotely close to observational data? The puzzle is\nusually resolved by subsuming the Milne Cosmological model into General\nRelativistic cosmology as the special case of an empty Universe. This\nexplanation may have to be reassessed with the finding that spacetime is\napproximately flat because of inflation, whereupon the projection of\ncosmological events onto the observer's Minkowski spacetime must always be\nkinematically consistent with Special Relativity, although the specific\ndynamics of the underlying General Relativistic model can give rise to virtual\nforces in order to maintain consistency between the observation and model\nframes.\n", " An inspiration at the origin of wavelet analysis (when Grossmann, Morlet,\nMeyer and collaborators were interacting and exploring versions of multiscale\nrepresentations) was provided by the analysis of holomorphic signals, for which\nthe images of the phase of Cauchy wavelets were remarkable in their ability to\nreveal intricate singularities or dynamic structures, such as instantaneous\nfrequency jumps in musical recordings. Our goal is to follow their seminal work\nand introduce recent developments in nonlinear analysis. In particular we\nsketch methods extending conventional Fourier analysis, exploiting both phase\nand amplitudes of holomorphic functions. The Blaschke factors are a key\ningredient, in building analytic tools, starting with the Malmquist Takenaka\northonormal bases of the Hardy space, continuing with \"best\" adapted bases\nobtained through phase unwinding, and concluding with relations to composition\nof Blaschke products and their dynamics. We also remark that the phase of a\nBlaschke product is a one layer neural net with arctan as an activation sigmoid\nand that the composition is a \"Deep Neural Net\" whose depth is the number of\ncompositions. Our results provide a wealth of related library of orthonormal\nbases.\n", " Topological methods, including persistent homology, are powerful tools for\nanalysis of high-dimensional data sets but these methods rely almost\nexclusively on thresholding techniques. In noisy data sets, thresholding does\nnot always allow for the recovery of topological information. We present an\neasy to implement, computationally efficient pre-processing algorithm to\nprepare noisy point cloud data sets for topological data analysis. The\ntopological de-noising algorithm allows for the recovery of topological\ninformation that is inaccessible by thresholding methods. We apply the\nalgorithm to synthetically-generated noisy data sets and show the recovery of\ntopological information which is impossible to obtain by thresholding. We also\napply the algorithm to natural image data in R^8 and show a very clean recovery\nof topological information previously only available with large amounts of\nthresholding. Finally, we discuss future directions for improving this\nalgorithm using zig-zag persistence methods.\n", " Normal estimation on 3D point clouds is a fundamental problem in 3D vision\nand graphics. Current methods often show limited accuracy in predicting normals\nat sharp features (e.g., edges and corners) and less robustness to noise. In\nthis paper, we propose a novel normal estimation method for point clouds. It\nconsists of two phases: (a) feature encoding which learns representations of\nlocal patches, and (b) normal estimation that takes the learned representation\nas input and regresses the normal vector. We are motivated that local patches\non isotropic and anisotropic surfaces have similar or distinct normals, and\nthat separable features or representations can be learned to facilitate normal\nestimation. To realise this, we first construct triplets of local patches on 3D\npoint cloud data, and design a triplet network with a triplet loss for feature\nencoding. We then design a simple network with several MLPs and a loss function\nto regress the normal vector. Despite having a smaller network size compared to\nmost other methods, experimental results show that our method preserves sharp\nfeatures and achieves better normal estimation results on CAD-like shapes.\n", " Recent work has demonstrated that the interior material layout of a 3D model\ncan be designed to make a fabricated replica satisfy application-specific\ndemands on its physical properties such as resistance to external loads. A\nwidely used practice to fabricate such models is by layer-based additive\nmanufacturing (AM) or 3D printing, which however suffers from the problem of\nadding and removing interior supporting structures. In this paper, we present a\nnovel method for generating support-free elliptic hollowing for 3D shapes which\ncan entirely avoid additional supporting structures. To achieve this, we\nperform the ellipse hollowing in the polygons on parallel section planes and\nprotrude the ellipses of one plane to its neighboring planes. To pack the\nellipses in a polygon, we construct the Voronoi diagram of ellipses to\nefficiently reason the free-space around the ellipses and other geometric\nfeatures by taking advantage of the available algorithm for the efficient and\nrobust construction of the Voronoi diagram of circles. We demonstrate the\neffectiveness and feasibility of our proposed method by generating interior\ndesigns for and printing various 3D shapes.\n", " This paper is devoted to study cylindrically symmetric stellar filaments in\nself-interacting Brans-Dicke gravity. For this purpose, we construct polytropic\nfilamentary models through generalized Lane-Emden equation in Newtonian regime.\nThe resulting models depend upon the values of cosmological constant (due to\nscalar field) along with polytropic index and represent a generalization of the\ncorresponding models in general relativity. We also investigate fragmentation\nof filaments by exploring the radial oscillations through stability analysis.\nThis stability criteria depends only upon the adiabatic index.\n" ]

[ " Bayesian inference provides a rigorous methodology for estimation and\nuncertainty quantification of parameters in geophysical forward models.\nBadlands (basin and landscape dynamics model) is a landscape evolution model\nthat simulates topography development at various space and time scales.\nBadlands consists of a number of geophysical parameters that needs estimation\nwith appropriate uncertainty quantification; given the observed present-day\nground truth such as surface topography and the stratigraphy of sediment\ndeposition through time. The inference of unknown parameters is challenging due\nto the scarcity of data, sensitivity of the parameter setting and complexity of\nthe Badlands model. In this paper, we take a Bayesian approach to provide\ninference using Markov chain Monte Carlo sampling (MCMC). We present\n\\textit{Bayeslands}; a Bayesian framework for Badlands that fuses information\nobtained from complex forward models with observational data and prior\nknowledge. As a proof-of-concept, we consider a synthetic and real-world\ntopography with two parameters for Bayeslands inference, namely precipitation\nand erodibility. The results of the experiments show that Bayeslands yields a\npromising distribution of the parameters. Moreover, we demonstrate the\nchallenge in sampling irregular and multi-modal posterior distributions using a\nlikelihood surface that has a range of sub-optimal modes.\n", " Attention-based sequence-to-sequence model has proved successful in Neural\nMachine Translation (NMT). However, the attention without consideration of\ndecoding history, which includes the past information in the decoder and the\nattention mechanism, often causes much repetition. To address this problem, we\npropose the decoding-history-based Adaptive Control of Attention (ACA) for the\nNMT model. ACA learns to control the attention by keeping track of the decoding\nhistory and the current information with a memory vector, so that the model can\ntake the translated contents and the current information into consideration.\nExperiments on Chinese-English translation and the English-Vietnamese\ntranslation have demonstrated that our model significantly outperforms the\nstrong baselines. The analysis shows that our model is capable of generating\ntranslation with less repetition and higher accuracy. The code will be\navailable at https://github.com/lancopku\n", " Learning the kernel parameters for Gaussian processes is often the\ncomputational bottleneck in applications such as online learning, Bayesian\noptimization, or active learning. Amortizing parameter inference over different\ndatasets is a promising approach to dramatically speed up training time.\nHowever, existing methods restrict the amortized inference procedure to a fixed\nkernel structure. The amortization network must be redesigned manually and\ntrained again in case a different kernel is employed, which leads to a large\noverhead in design time and training time. We propose amortizing kernel\nparameter inference over a complete kernel-structure-family rather than a fixed\nkernel structure. We do that via defining an amortization network over pairs of\ndatasets and kernel structures. This enables fast kernel inference for each\nelement in the kernel family without retraining the amortization network. As a\nby-product, our amortization network is able to do fast ensembling over kernel\nstructures. In our experiments, we show drastically reduced inference time\ncombined with competitive test performance for a large set of kernels and\ndatasets.\n", " This paper deals with the modular irregularity strength of a graph of n\nvertices, a new graph invariant, modified from the irregularity strength, by\nchanging the condition of the vertex-weight set associate to the well-known\nirregular labeling from n distinct positive integer to Z_n-the group of integer\nmodulo n. Investigating the triangular book graph B_m^((3)), we first find the\nirregularity strength of triangular book graph s(B_m^((3)) ), as the lower\nbound for the modular irregularity strength, and then construct a modular\nirregular s(B_m^((3)) )-labeling. The result shows that triangular book graphs\nadmit a modular irregular labeling and its modular irregularity strength and\nirregularity strength are equal, except for a small case and the infinity\nproperty.\n", " We perform a detailed study of a class of irregular correlators in Liouville\nConformal Field Theory, of the related Virasoro conformal blocks with irregular\nsingularities and of their connection formulae. Upon considering their\nsemi-classical limit, we provide explicit expressions of the connection\nmatrices for the Heun function and a class of its confluences. Their\ncalculation is reduced to concrete combinatorial formulae from conformal block\nexpansions.\n", " A new large-scale parallel multiconfigurational self-consistent field (MCSCF)\nimplementation in the open-source NWChem computational chemistry code is\npresented. The generalized active space (GAS) approach is used to partition\nlarge configuration interaction (CI) vectors and generate a sufficient number\nof batches that can be distributed to the available nodes. Massively parallel\nCI calculations with large active spaces can be treated. The performance of the\nnew parallel MCSCF implementation is presented for the chromium trimer and for\nan active space of 20 electrons in 20 orbitals. Unprecedented CI calculations\nwith an active space of 22 electrons in 22 orbitals for the pentacene systems\nwere performed and a single CI iteration calculation with an active space of 24\nelectrons in 24 orbitals for the chromium tetramer was possible. The chromium\ntetramer corresponds to a CI expansion of one trillion SDs (914 058 513 424)\nand is largest conventional CI calculation attempted up to date.\n", " In this paper, we prove that for $s\\in(1,2)$ there exists no totally lower\nirregular finite positive Borel measure $\\mu$ in $\\R^2$ with\\break $\\mathcal\nH^s(\\supp\\mu)<+\\infty$ such that $\\|R\\mu\\|\\ci{L^\\infty(m_2)}<+\\infty$, where\n$R\\mu=\\mu\\ast\\frac{x}{|x|^{s+1}}$ and $m_2$ is the Lebesgue measure in $\\R^2$.\nCombined with known results of Prat and Vihtil\\\"a, this shows that for any\nnon-integer $s\\in(0,2)$ and any finite positive Borel measure in $\\R^2$ with\n$\\mathcal H^s(\\supp\\mu)<+\\infty$, we have $\\|R\\mu\\|\\ci{L^\\infty(m_2)}=\\infty$.\n", " We study the relationship between stable sampling sequences for bandlimited\nfunctions in $L^p(\\R^n)$ and the Fourier multipliers in $L^p$. In the case that\nthe sequence is a lattice and the spectrum is a fundamental domain for the\nlattice the connection is complete. In the case of irregular sequences there is\nstill a partial relationship.\n", " Network Pruning is a promising way to address the huge computing resource\ndemands of the deployment and inference of Large Language Models (LLMs).\nRetraining-free is important for LLMs' pruning methods. However, almost all of\nthe existing retraining-free pruning approaches for LLMs focus on unstructured\npruning, which requires specific hardware support for acceleration. In this\npaper, we propose a novel retraining-free structured pruning framework for\nLLMs, named FLAP (FLuctuation-based Adaptive Structured Pruning). It is\nhardware-friendly by effectively reducing storage and enhancing inference\nspeed. For effective structured pruning of LLMs, we highlight three critical\nelements that demand the utmost attention: formulating structured importance\nmetrics, adaptively searching the global compressed model, and implementing\ncompensation mechanisms to mitigate performance loss. First, FLAP determines\nwhether the output feature map is easily recoverable when a column of weight is\nremoved, based on the fluctuation pruning metric. Then it standardizes the\nimportance scores to adaptively determine the global compressed model\nstructure. At last, FLAP adds additional bias terms to recover the output\nfeature maps using the baseline values. We thoroughly evaluate our approach on\na variety of language benchmarks. Without any retraining, our method\nsignificantly outperforms the state-of-the-art methods, including LLM-Pruner\nand the extension of Wanda in structured pruning. The code is released at\nhttps://github.com/CASIA-IVA-Lab/FLAP.\n", " We propose a novel molecular computing scheme for statistical inference. We\nfocus on the much-studied statistical inference problem of computing maximum\nlikelihood estimators for log-linear models. Our scheme takes log-linear models\nto reaction systems, and the observed data to initial conditions, so that the\ncorresponding equilibrium of each reaction system encodes the corresponding\nmaximum likelihood estimator. The main idea is to exploit the coincidence\nbetween thermodynamic entropy and statistical entropy. We map a Maximum Entropy\ncharacterization of the maximum likelihood estimator onto a Maximum Entropy\ncharacterization of the equilibrium concentrations for the reaction system.\nThis allows for an efficient encoding of the problem, and reveals that reaction\nnetworks are superbly suited to statistical inference tasks. Such a scheme may\nalso provide a template to understanding how in vivo biochemical signaling\npathways integrate extensive information about their environment and history.\n", " This paper addresses the problem of statistical inference for latent\ncontinuous-time stochastic processes, which is often intractable, particularly\nfor discrete state space processes described by Markov jump processes. To\novercome this issue, we propose a new tractable inference scheme based on an\nentropic matching framework that can be embedded into the well-known\nexpectation propagation algorithm. We demonstrate the effectiveness of our\nmethod by providing closed-form results for a simple family of approximate\ndistributions and apply it to the general class of chemical reaction networks,\nwhich are a crucial tool for modeling in systems biology. Moreover, we derive\nclosed form expressions for point estimation of the underlying parameters using\nan approximate expectation maximization procedure. We evaluate the performance\nof our method on various chemical reaction network instantiations, including a\nstochastic Lotka-Voltera example, and discuss its limitations and potential for\nfuture improvements. Our proposed approach provides a promising direction for\naddressing complex continuous-time Bayesian inference problems.\n", " In this paper we present a rare combination of abstract results on the\nspectral properties of slanted matrices and some of their very specific\napplications to frame theory and sampling problems. We show that for a large\nclass of slanted matrices boundedness below of the corresponding operator in\n$\\ell^p$ for some $p$ implies boundedness below in $\\ell^p$ for all $p$. We use\nthe established resultto enrich our understanding of Banach frames and obtain\nnew results for irregular sampling problems. We also present a version of a\nnon-commutative Wiener's lemma for slanted matrices.\n", " Variational inference is a popular technique to approximate a possibly\nintractable Bayesian posterior with a more tractable one. Recently, boosting\nvariational inference has been proposed as a new paradigm to approximate the\nposterior by a mixture of densities by greedily adding components to the\nmixture. However, as is the case with many other variational inference\nalgorithms, its theoretical properties have not been studied. In the present\nwork, we study the convergence properties of this approach from a modern\noptimization viewpoint by establishing connections to the classic Frank-Wolfe\nalgorithm. Our analyses yields novel theoretical insights regarding the\nsufficient conditions for convergence, explicit rates, and algorithmic\nsimplifications. Since a lot of focus in previous works for variational\ninference has been on tractability, our work is especially important as a much\nneeded attempt to bridge the gap between probabilistic models and their\ncorresponding theoretical properties.\n", " Graph Neural Networks (GNNs) have emerged as a powerful and flexible\nframework for representation learning on irregular data. As they generalize the\noperations of classical CNNs on grids to arbitrary topologies, GNNs also bring\nmuch of the implementation challenges of their Euclidean counterparts. Model\nsize, memory footprint, and energy consumption are common concerns for many\nreal-world applications. Network binarization allocates a single bit to\nparameters and activations, thus dramatically reducing the memory requirements\n(up to 32x compared to single-precision floating-point numbers) and maximizing\nthe benefits of fast SIMD instructions on modern hardware for measurable\nspeedups. However, in spite of the large body of work on binarization for\nclassical CNNs, this area remains largely unexplored in geometric deep\nlearning. In this paper, we present and evaluate different strategies for the\nbinarization of graph neural networks. We show that through careful design of\nthe models, and control of the training process, binary graph neural networks\ncan be trained at only a moderate cost in accuracy on challenging benchmarks.\nIn particular, we present the first dynamic graph neural network in Hamming\nspace, able to leverage efficient k-NN search on binary vectors to speed-up the\nconstruction of the dynamic graph. We further verify that the binary models\noffer significant savings on embedded devices. Our code is publicly available\non Github.\n", " Over the years, scene understanding has attracted a growing interest in\ncomputer vision, providing the semantic and physical scene information\nnecessary for robots to complete some particular tasks autonomously. In 3D\nscenes, rich spatial geometric and topological information are often ignored by\nRGB-based approaches for scene understanding. In this study, we develop a\nbottom-up approach for scene understanding that infers support relations\nbetween objects from a point cloud. Our approach utilizes the spatial topology\ninformation of the plane pairs in the scene, consisting of three major steps.\n1) Detection of pairwise spatial configuration: dividing primitive pairs into\nlocal support connection and local inner connection; 2) primitive\nclassification: a combinatorial optimization method applied to classify\nprimitives; and 3) support relations inference and hierarchy graph\nconstruction: bottom-up support relations inference and scene hierarchy graph\nconstruction containing primitive level and object level. Through experiments,\nwe demonstrate that the algorithm achieves excellent performance in primitive\nclassification and support relations inference. Additionally, we show that the\nscene hierarchy graph contains rich geometric and topological information of\nobjects, and it possesses great scalability for scene understanding.\n", " Recently, Kitaev [Ki2] introduced partially ordered generalized patterns\n(POGPs) in the symmetric group, which further generalize the generalized\npermutation patterns introduced by Babson and Steingr\\'imsson [BS]. A POGP p is\na GP some of whose letters are incomparable. In this paper, we study the\ngenerating functions (g.f.) for the number of k-ary words avoiding some POGPs.\nWe give analogues, extend and generalize several known results, as well as get\nsome new results. In particular, we give the g.f. for the entire distribution\nof the maximum number of non-overlapping occurrences of a pattern p with no\nhyphens (that allowed to have repetition of letters), provided we know the g.f.\nfor the number of k-ary words that avoid p.\n", " Protein phosphorylation is a reversible post-translational modification\ncommonly used by cell signaling networks to transmit information about the\nextracellular environment into intracellular organelles for the regulation of\nthe activity and sorting of proteins within the cell. For this study we\nreconstructed a literature-based mammalian kinase-substrate network from\nseveral online resources. The interactions within this directed graph network\nconnect kinases to their substrates, through specific phosphosites including\nkinase-kinase regulatory interactions. However, the \"signs\" of links,\nactivation or inhibition of the substrate upon phosphorylation, within this\nnetwork are mostly unknown. Here we show how we can infer the \"signs\"\nindirectly using data from quantitative phosphoproteomics experiments applied\nto mammalian cells combined with the literature-based kinase-substrate network.\nOur inference method was able to predict the sign for 321 links and 153\nphosphosites on 120 kinases, resulting in signed and directed subnetwork of\nmammalian kinase-kinase interactions. Such an approach can rapidly advance the\nreconstruction of cell signaling pathways and networks regulating mammalian\ncells.\n", " Neural codes serve as a language for neurons in the brain. Convex codes,\nwhich arise from the pattern of intersections of convex sets in Euclidean\nspace, are of particular relevance to neuroscience. Not every code is convex,\nhowever, and the combinatorial properties of a code that determine its\nconvexity are still poorly understood. Here we find that a code that can be\nrealized by a collection of open convex sets may or may not be realizable by\nclosed convex sets, and vice versa, establishing that open convex and closed\nconvex codes are distinct classes. We also prove that max intersection-complete\ncodes (i.e. codes that contain all intersections of maximal codewords) are both\nopen convex and closed convex, and provide an upper bound for their minimal\nembedding dimension. Finally, we show that the addition of non-maximal\ncodewords to an open convex code preserves convexity.\n", " Several adaptation techniques have been investigated to optimize fuzzy\ninference systems. Neural network learning algorithms have been used to\ndetermine the parameters of fuzzy inference system. Such models are often\ncalled as integrated neuro-fuzzy models. In an integrated neuro-fuzzy model\nthere is no guarantee that the neural network learning algorithm converges and\nthe tuning of fuzzy inference system will be successful. Success of\nevolutionary search procedures for optimization of fuzzy inference system is\nwell proven and established in many application areas. In this paper, we will\nexplore how the optimization of fuzzy inference systems could be further\nimproved using a meta-heuristic approach combining neural network learning and\nevolutionary computation. The proposed technique could be considered as a\nmethodology to integrate neural networks, fuzzy inference systems and\nevolutionary search procedures. We present the theoretical frameworks and some\nexperimental results to demonstrate the efficiency of the proposed technique.\n", " Vertical federated learning (VFL), where data features are stored in multiple\nparties distributively, is an important area in machine learning. However, the\ncommunication complexity for VFL is typically very high. In this paper, we\npropose a unified framework by constructing coresets in a distributed fashion\nfor communication-efficient VFL. We study two important learning tasks in the\nVFL setting: regularized linear regression and $k$-means clustering, and apply\nour coreset framework to both problems. We theoretically show that using\ncoresets can drastically alleviate the communication complexity, while nearly\nmaintain the solution quality. Numerical experiments are conducted to\ncorroborate our theoretical findings.\n", " We give some additions to the article \"On the generalized Riemann-Hilbert\nproblem with irregular singularities\" by Bolibruch, Malek, Mitschi\n(math/0410483). In particular, a weak GRH-problem and the GRH-problem for\nscalar differential equations are discussed.\n", " We introduce an unsupervised clustering algorithm to improve training\nefficiency and accuracy in predicting energies using molecular-orbital-based\nmachine learning (MOB-ML). This work determines clusters via the Gaussian\nmixture model (GMM) in an entirely automatic manner and simplifies an earlier\nsupervised clustering approach [J. Chem. Theory Comput., 15, 6668 (2019)] by\neliminating both the necessity for user-specified parameters and the training\nof an additional classifier. Unsupervised clustering results from GMM have the\nadvantage of accurately reproducing chemically intuitive groupings of frontier\nmolecular orbitals and having improved performance with an increasing number of\ntraining examples. The resulting clusters from supervised or unsupervised\nclustering is further combined with scalable Gaussian process regression (GPR)\nor linear regression (LR) to learn molecular energies accurately by generating\na local regression model in each cluster. Among all four combinations of\nregressors and clustering methods, GMM combined with scalable exact Gaussian\nprocess regression (GMM/GPR) is the most efficient training protocol for\nMOB-ML. The numerical tests of molecular energy learning on thermalized\ndatasets of drug-like molecules demonstrate the improved accuracy,\ntransferability, and learning efficiency of GMM/GPR over not only other\ntraining protocols for MOB-ML, i.e., supervised regression-clustering combined\nwith GPR(RC/GPR) and GPR without clustering. GMM/GPR also provide the best\nmolecular energy predictions compared with the ones from literature on the same\nbenchmark datasets. With a lower scaling, GMM/GPR has a 10.4-fold speedup in\nwall-clock training time compared with scalable exact GPR with a training size\nof 6500 QM7b-T molecules.\n", " We propose an end-to-end pipeline to robustly generate high-quality\nquadrilateral meshes for complex CAD models. An initial quad-dominant mesh is\ngenerated with frontal point insertion guided by a locally integrable cross\nfield and a scalar size map adapted to the small CAD features. After triangle\ncombination and midpoint-subdivision into an all-quadrilateral mesh, the\ntopology of the mesh is modified to reduce the number of irregular vertices.\nThe idea is to preserve the irregular vertices matching cross-field\nsingularities and to eliminate the others. The topological modifications are\neither local and based on disk quadrangulations, or more global with the\nremeshing of patches of quads according to predefined patterns. Validity of the\nquad mesh is guaranteed by monitoring element quality during all operations and\nreverting the changes when necessary. Advantages of our approach include\nrobustness, strict respect of the CAD features and support for user-prescribed\nsize constraints. The quad mesher, which is available in Gmsh, is validated and\nillustrated on two datasets of CAD models.\n", " Richard Cox [1] set the axiomatic foundations of probable inference and the\nalgebra of propositions. He showed that consistency within these axioms\nrequires certain rules for updating belief. In this paper we use the analogy\nbetween probability and utility introduced in [2] to propose an axiomatic\nfoundation for utility inference and the algebra of preferences. We show that\nconsistency within these axioms requires certain rules for updating preference.\nWe discuss a class of utility functions that stems from the axioms of utility\ninference and show that this class is the basic building block for any general\nmultiattribute utility function. We use this class of utility functions\ntogether with the algebra of preferences to construct utility functions\nrepresented by logical operations on the attributes.\n", " Although Ornstein's nonestimate entails the impossibility to control in\ngeneral all the $L^1$-norm of derivatives of a function by the $L^1$-norm of a\nconstant coefficient homogeneous vector differential operator, the\ncorresponding endpoint Sobolev inequality has been known to hold in many cases:\nthe gradient of scalar functions (Gagliardo and Nirenberg), the deformation\noperator (Korn-Sobolev inequality by M.J. Strauss), and the Hodge complex\n(Bourgain and Brezis). The class of differential operators for which estimates\nholds can be characterized by a cancelling condition. The proof of the\nestimates rely on a duality estimate for $L^1$-vector fields lying in the\nkernel of a cocancelling differential operator, combined with classical linear\nalgebra and harmonic analysis techniques. This characterization unifies classes\nof known Sobolev inequalities and extends to fractional Sobolev and Hardy\ninequalities. A similar weaker condition introduced by Rai\\c{t}\\u{a}\ncharacterizes the operators for which there is an $L^\\infty$-estimate on\nlower-order derivatives.\n", " In diffusion MRI (dMRI), a good sampling scheme is important for efficient\nacquisition and robust reconstruction. Diffusion weighted signal is normally\nacquired on single or multiple shells in q-space. Signal samples are typically\ndistributed uniformly on different shells to make them invariant to the\norientation of structures within tissue, or the laboratory coordinate frame.\nThe Electrostatic Energy Minimization (EEM) method, originally proposed for\nsingle shell sampling scheme in dMRI, was recently generalized to multi-shell\nschemes, called Generalized EEM (GEEM). GEEM has been successfully used in the\nHuman Connectome Project (HCP). However, EEM does not directly address the goal\nof optimal sampling, i.e., achieving large angular separation between sampling\npoints. In this paper, we propose a more natural formulation, called Spherical\nCode (SC), to directly maximize the minimal angle between different samples in\nsingle or multiple shells. We consider not only continuous problems to design\nsingle or multiple shell sampling schemes, but also discrete problems to\nuniformly extract sub-sampled schemes from an existing single or multiple shell\nscheme, and to order samples in an existing scheme. We propose five algorithms\nto solve the above problems, including an incremental SC (ISC), a sophisticated\ngreedy algorithm called Iterative Maximum Overlap Construction (IMOC), an 1-Opt\ngreedy method, a Mixed Integer Linear Programming (MILP) method, and a\nConstrained Non-Linear Optimization (CNLO) method. To our knowledge, this is\nthe first work to use the SC formulation for single or multiple shell sampling\nschemes in dMRI. Experimental results indicate that SC methods obtain larger\nangular separation and better rotational invariance than the state-of-the-art\nEEM and GEEM. The related codes and a tutorial have been released in DMRITool.\n", " A data structure is presented for point location in connected planar\nsubdivisions when the distribution of queries is known in advance. The data\nstructure has an expected query time that is within a constant factor of\noptimal. More specifically, an algorithm is presented that preprocesses a\nconnected planar subdivision G of size n and a query distribution D to produce\na point location data structure for G. The expected number of point-line\ncomparisons performed by this data structure, when the queries are distributed\naccording to D, is H + O(H^{2/3}+1) where H=H(G,D) is a lower bound on the\nexpected number of point-line comparisons performed by any linear decision tree\nfor point location in G under the query distribution D. The preprocessing\nalgorithm runs in O(n log n) time and produces a data structure of size O(n).\nThese results are obtained by creating a Steiner triangulation of G that has\nnear-minimum entropy.\n", " We introduce the problem of finding a set $B$ of $k$ points in $[0,1]^n$ such\nthat the expected cost of the cheapest point in $B$ that dominates a random\npoint from $[0,1]^n$ is minimized. We study the case where the coordinates of\nthe random points are independently distributed and the cost function is\nlinear. This problem arises naturally in various application areas where\ncustomers' requests are satisfied based on predefined products, each\ncorresponding to a subset of features. We show that the problem is NP-hard\nalready for $k=2$ when each coordinate is drawn from $\\{0,1\\}$, and obtain an\nFPTAS for general fixed $k$ under mild assumptions on the distributions.\n", " Triangulations are an ubiquitous input for the finite element community.\nHowever, most raw triangulations obtained by imaging techniques are unsuitable\nas-is for finite element analysis. In this paper, we give a robust pipeline for\nhandling those triangulations, based on the computation of a one-to-one\nparametrization for automatically selected patches of input triangles, which\nmakes each patch amenable to remeshing by standard finite element meshing\nalgorithms. Using only geometrical arguments, we prove that a discrete\nparametrization of a patch is one-to-one if (and only if) its image in the\nparameter space is such that all parametric triangles have a positive area. We\nthen derive a non-standard linear discretization scheme based on mean value\ncoordinates to compute such one-to-one parametrizations, and show that the\nscheme does not discretize a Laplacian on a structured mesh. The proposed\npipeline is implemented in the open source mesh generator Gmsh, where the\ncreation of suitable patches is based on triangulation topology and\nparametrization quality, combined with feature edge detection. Several examples\nillustrate the robustness of the resulting implementation.\n", " In this paper, we present an end-to-end joint entity and relation extraction\napproach based on transformer-based language models. We apply the model to the\ntask of linking mathematical symbols to their descriptions in LaTeX documents.\nIn contrast to existing approaches, which perform entity and relation\nextraction in sequence, our system incorporates information from relation\nextraction into entity extraction. This means that the system can be trained\neven on data sets where only a subset of all valid entity spans is annotated.\nWe provide an extensive evaluation of the proposed system and its strengths and\nweaknesses. Our approach, which can be scaled dynamically in computational\ncomplexity at inference time, produces predictions with high precision and\nreaches 3rd place in the leaderboard of SemEval-2022 Task 12. For inputs in the\ndomain of physics and math, it achieves high relation extraction macro F1\nscores of 95.43% and 79.17%, respectively. The code used for training and\nevaluating our models is available at: https://github.com/nicpopovic/RE1st\n", " A coreset (or core-set) of an input set is its small summation, such that\nsolving a problem on the coreset as its input, provably yields the same result\nas solving the same problem on the original (full) set, for a given family of\nproblems (models, classifiers, loss functions). Over the past decade, coreset\nconstruction algorithms have been suggested for many fundamental problems in\ne.g. machine/deep learning, computer vision, graphics, databases, and\ntheoretical computer science. This introductory paper was written following\nrequests from (usually non-expert, but also colleagues) regarding the many\ninconsistent coreset definitions, lack of available source code, the required\ndeep theoretical background from different fields, and the dense papers that\nmake it hard for beginners to apply coresets and develop new ones.\n The paper provides folklore, classic and simple results including\nstep-by-step proofs and figures, for the simplest (accurate) coresets of very\nbasic problems, such as: sum of vectors, minimum enclosing ball, SVD/ PCA and\nlinear regression. Nevertheless, we did not find most of their constructions in\nthe literature. Moreover, we expect that putting them together in a\nretrospective context would help the reader to grasp modern results that\nusually extend and generalize these fundamental observations. Experts might\nappreciate the unified notation and comparison table that links between\nexisting results.\n Open source code with example scripts are provided for all the presented\nalgorithms, to demonstrate their practical usage, and to support the readers\nwho are more familiar with programming than math.\n", " We consider fractional differential equations of order $\\alpha \\in (0,1)$ for\nfunctions of one independent variable $t\\in (0,\\infty)$ with the\nRiemann-Liouville and Caputo-Dzhrbashyan fractional derivatives. A precise\nestimate for the order of growth of $\\alpha$-entire solutions is given. An\nanalog of the Frobenius method for systems with regular singularity is\ndeveloped. For a model example of an equation with a kind of an irregular\nsingularity, a series for a formal solution is shown to be convergent for $t>0$\n(if $\\alpha$ is an irrational number poorly approximated by rational ones) but\ndivergent in the distribution sense.\n", " Grid computing is distributed computing performed transparently across\nmultiple administrative domains. Grid middleware, which is meant to enable\naccess to grid resources, is currently widely seen as being too heavyweight\nand, in consequence, unwieldy for general scientific use. Its heavyweight\nnature, especially on the client-side, has severely restricted the uptake of\ngrid technology by computational scientists. In this paper, we describe the\nApplication Hosting Environment (AHE) which we have developed to address some\nof these problems. The AHE is a lightweight, easily deployable environment\ndesigned to allow the scientist to quickly and easily run legacy applications\non distributed grid resources. It provides a higher level abstraction of a grid\nthan is offered by existing grid middleware schemes such as the Globus Toolkit.\nAs a result the computational scientist does not need to know the details of\nany particular underlying grid middleware and is isolated from any changes to\nit on the distributed resources. The functionality provided by the AHE is\n`application-centric': applications are exposed as web services with a\nwell-defined standards-compliant interface. This allows the computational\nscientist to start and manage application instances on a grid in a transparent\nmanner, thus greatly simplifying the user experience. We describe how a range\nof computational science codes have been hosted within the AHE and how the\ndesign of the AHE allows us to implement complex workflows for deployment on\ngrid infrastructure.\n", " This paper introduces the concept of Bessel multipliers. These operators are\ndefined by a fixed multiplication pattern, which is inserted between the\nAnalysis and synthesis operators. The proposed concept unifies the approach\nused for Gabor multipliers for arbitrary analysis/synthesis systems, which form\nBessel sequences, like wavelet or irregular Gabor frames. The basic properties\nof this class of operators are investigated. In particular the implications of\nsummability properties of the symbol for the membership of the corresponding\noperators in certain operator classes are specified. As a special case the\nmultipliers for Riesz bases are examined and it is shown that multipliers in\nthis case can be easily composed and inverted. Finally the continuous\ndependence of a Bessel multiplier on the parameters (i.e. the involved\nsequences and the symbol in use) is verified, using a special measure of\nsimilarity of sequences.\n", " We propose a data-driven method for recovering miss-ing parts of 3D shapes.\nOur method is based on a new deep learning architecture consisting of two\nsub-networks: a global structure inference network and a local geometry\nrefinement network. The global structure inference network incorporates a long\nshort-term memorized context fusion module (LSTM-CF) that infers the global\nstructure of the shape based on multi-view depth information provided as part\nof the input. It also includes a 3D fully convolutional (3DFCN) module that\nfurther enriches the global structure representation according to volumetric\ninformation in the input. Under the guidance of the global structure network,\nthe local geometry refinement network takes as input lo-cal 3D patches around\nmissing regions, and progressively produces a high-resolution, complete surface\nthrough a volumetric encoder-decoder architecture. Our method jointly trains\nthe global structure inference and local geometry refinement networks in an\nend-to-end manner. We perform qualitative and quantitative evaluations on six\nobject categories, demonstrating that our method outperforms existing\nstate-of-the-art work on shape completion.\n", " We study linear abstract differential-algebraic equations (ADAEs), and we\nintroduce an index concept which is based on polynomial growth of\na~pseudo-resolvent. Our approach to solvability analysis is based on degenerate\nsemigroups. We apply our results to some examples such as distributed circuit\nelements, and a system obtained by heat-wave coupling.\n", " We present the public version of hi_class (www.hiclass-code.net), an\nextension of the Boltzmann code CLASS to a broad ensemble of modifications to\ngeneral relativity. In particular, hi_class can calculate predictions for\nmodels based on Horndeski's theory, which is the most general scalar-tensor\ntheory described by second-order equations of motion and encompasses any\nperfect-fluid dark energy, quintessence, Brans-Dicke, $f(R)$ and covariant\nGalileon models. hi_class has been thoroughly tested and can be readily used to\nunderstand the impact of alternative theories of gravity on linear structure\nformation as well as for cosmological parameter extraction.\n", " Motivated by data-rich experiments in transcriptional regulation and sensory\nneuroscience, we consider the following general problem in statistical\ninference. When exposed to a high-dimensional signal S, a system of interest\ncomputes a representation R of that signal which is then observed through a\nnoisy measurement M. From a large number of signals and measurements, we wish\nto infer the \"filter\" that maps S to R. However, the standard method for\nsolving such problems, likelihood-based inference, requires perfect a priori\nknowledge of the \"noise function\" mapping R to M. In practice such noise\nfunctions are usually known only approximately, if at all, and using an\nincorrect noise function will typically bias the inferred filter. Here we show\nthat, in the large data limit, this need for a pre-characterized noise function\ncan be circumvented by searching for filters that instead maximize the mutual\ninformation I[M;R] between observed measurements and predicted representations.\nMoreover, if the correct filter lies within the space of filters being\nexplored, maximizing mutual information becomes equivalent to simultaneously\nmaximizing every dependence measure that satisfies the Data Processing\nInequality. It is important to note that maximizing mutual information will\ntypically leave a small number of directions in parameter space unconstrained.\nWe term these directions \"diffeomorphic modes\" and present an equation that\nallows these modes to be derived systematically. The presence of diffeomorphic\nmodes reflects a fundamental and nontrivial substructure within parameter\nspace, one that is obscured by standard likelihood-based inference.\n", " It is known that the continuous wavelet transform of a function $f$ decays\nvery rapidly near the points where $f$ is smooth, while it decays slowly near\nthe irregular points. This property allows one to precisely identify the\nsingular support of $f$. However, the continuous wavelet transform is unable to\nprovide additional information about the geometry of the singular points. In\nthis paper, we introduce a new transform for functions and distributions on\n$\\R^2$, called the Continuous Shearlet Transform. This is defined by\n$\\mathcal{S}\\mathcal{H}_f(a,s,t) = \\ip{f}{\\psi_{ast}}$, where the analyzing\nelements $\\psi_{ast}$ are dilated and translated copies of a single generating\nfunction $\\psi$ and, thus, they form an affine system. The resulting continuous\nshearlets $\\psi_{ast}$ are smooth functions at continuous scales $a >0$,\nlocations $t \\in \\R^2$ and oriented along lines of slope $s \\in \\R$ in the\nfrequency domain. The Continuous Shearlet Transform transform is able to\nidentify not only the location of the singular points of a distribution $f$,\nbut also the orientation of their distributed singularities. As a result, we\ncan use this transform to exactly characterize the wavefront set of $f$.\n", " The availability of accurate localization is critical for multi-robot\nexploration strategies; noisy or inconsistent localization causes failure in\nmeeting exploration objectives. We aim to achieve high localization accuracy\nwith contemporary exploration map belief and vice versa without needing global\nlocalization information. This paper proposes a novel simultaneous exploration\nand localization (SEAL) approach, which uses Gaussian Processes (GP)-based\ninformation fusion for maximum exploration while performing communication graph\noptimization for relative localization. Both these cross-dependent objectives\nwere integrated through the Rao-Blackwellization technique. Distributed\nlinearized convex hull optimization is used to select the next-best unexplored\nregion for distributed exploration. SEAL outperformed cutting-edge methods on\nexploration and localization performance in extensive ROS-Gazebo simulations,\nillustrating the practicality of the approach in real-world applications.\n", " We geometrically analyze the problem of estimating parameters related to the\nshape and size of a two-dimensional target object on the plane by using\nrandomly distributed distance sensors whose locations are unknown. Based on the\nanalysis using geometric probability, we discuss the observability of these\nparameters: which parameters we can estimate and what conditions are required\nto estimate them. For a convex target object, its size and perimeter length are\nobservable, and other parameters are not observable. For a general polygon\ntarget object, convexity in addition to its size and perimeter length is\nobservable. Parameters related to a concave vertex can be observable when some\nconditions are satisfied. We also propose a method for estimating the convexity\nof a target object and the perimeter length of the target object.\n", " In recent years, JavaScript has become the most widely used programming\nlanguage, especially in web development. However, writing secure JavaScript\ncode is not trivial, and programmers often make mistakes that lead to security\nvulnerabilities in web applications. Large Language Models (LLMs) have\ndemonstrated substantial advancements across multiple domains, and their\nevolving capabilities indicate their potential for automatic code generation\nbased on a required specification, including automatic bug fixing. In this\nstudy, we explore the accuracy of LLMs, namely ChatGPT and Bard, in finding and\nfixing security vulnerabilities in JavaScript programs. We also investigate the\nimpact of context in a prompt on directing LLMs to produce a correct patch of\nvulnerable JavaScript code. Our experiments on real-world software\nvulnerabilities show that while LLMs are promising in automatic program repair\nof JavaScript code, achieving a correct bug fix often requires an appropriate\namount of context in the prompt.\n", " This paper introduces Whittemore, a language for causal programming. Causal\nprogramming is based on the theory of structural causal models and consists of\ntwo primary operations: identification, which finds formulas that compute\ncausal queries, and estimation, which applies formulas to transform probability\ndistributions to other probability distribution. Causal programming provides\nabstractions to declare models, queries, and distributions with syntax similar\nto standard mathematical notation, and conducts rigorous causal inference,\nwithout requiring detailed knowledge of the underlying algorithms. Examples of\ncausal inference with real data are provided, along with discussion of the\nimplementation and possibilities for future extension.\n", " In the realm of deep learning, the self-attention mechanism has substantiated\nits pivotal role across a myriad of tasks, encompassing natural language\nprocessing and computer vision. Despite achieving success across diverse\napplications, the traditional self-attention mechanism primarily leverages\nlinear transformations for the computation of query, key, and value (QKV),\nwhich may not invariably be the optimal choice under specific circumstances.\nThis paper probes into a novel methodology for QKV computation-implementing a\nspecially-designed neural network structure for the calculation. Utilizing a\nmodified Marian model, we conducted experiments on the IWSLT 2017\nGerman-English translation task dataset and juxtaposed our method with the\nconventional approach. The experimental results unveil a significant\nenhancement in BLEU scores with our method. Furthermore, our approach also\nmanifested superiority when training the Roberta model with the Wikitext-103\ndataset, reflecting a notable reduction in model perplexity compared to its\noriginal counterpart. These experimental outcomes not only validate the\nefficacy of our method but also reveal the immense potential in optimizing the\nself-attention mechanism through neural network-based QKV computation, paving\nthe way for future research and practical applications. The source code and\nimplementation details for our proposed method can be accessed at\nhttps://github.com/ocislyjrti/NeuralAttention.\n", " Huffman Codes are optimal Instantaneous Fixed-to-Variable (FV) codes in which\nevery source symbol can only be encoded by one codeword. Relaxing these\nconstraints permits constructing better FV codes. More specifically, recent\nwork has shown that AIFV-$m$ codes can beat Huffman coding. AIFV-$m$ codes\nconstruct am $m$-tuple of different coding trees between which the code\nalternates and are only almost instantaneous (AI). This means that decoding a\nword might require a delay of a finite number of bits.\n Current algorithms for constructing optimal AIFV-$m$ codes are iterative\nprocesses that construct progressively \"better sets\" of code trees. The\nprocesses have been proven to finitely converge to the optimal code but with no\nknown bounds on the convergence rate.\n This paper derives a geometric interpretation of the space of binary AIFV-$2$\ncodes, permitting the development of the first polynomially time-bounded\niterative procedures for constructing optimal AIFV codes.\n We first show that a simple binary search procedure can replace the current\niterative process to construct optimal binary AIFV-$2$ codes. We then describe\nhow to frame the problem as a linear programming with an exponential number of\nconstraints but a polynomial-time separability oracle. This permits using the\nGr\\\"otschel, Lov\\'asz and Schrijver ellipsoid method to solve the problem in a\npolynomial number of steps. While more complicated, this second method has the\npotential to lead to a polynomial time algorithm to construct optimal AIFV-$m$\ncodes for general $m$.\n", " In the last couple of years there were a few attempts to apply topological\ndata analysis to text, and in particular to natural language inference. A\nrecent work by Tymochko et al. suggests the possibility of capturing `the\nnotion of logical shape in text,' using `topological delay embeddings,' a\ntechnique derived from dynamical systems, applied to word embeddings.\n In this note we reconstruct their argument and show, using several old and\nnew examples, that the problem of connecting logic, topology and text is still\nvery much unsolved. We conclude that there is no clear answer to the question:\n``Can we find a circle in a circular argument?'' We point out some possible\navenues of exploration. The code used in our experiment is also shown.\n", " A method of analyzing time bounds for randomized distributed algorithms is\npresented, in the context of a new and general framework for describing and\nreasoning about randomized algorithms. The method consists of proving auxiliary\nstatements of the form U (t)->(p) U', which means that whenever the algorithm\nbegins in a state in set U, with probability p, it will reach a state in set U'\nwithin time t. The power of the method is illustrated by its use in proving a\nconstant upper bound on the expected time for some process to reach its\ncritical region, in Lehmann and Rabin's Dining Philosophers algorithm.\n", " We show that a minor variant of the continuous Fr\\'echet distance between\npolygonal curves can be computed using essentially the same algorithm used to\nsolve the discrete version, thus dramatically simplifying the algorithm for\ncomputing it. The new variant is not necessarily monotone, but this shortcoming\ncan be easily handled via refinement.\n Combined with a Dijkstra/Prim type algorithm, this leads to a realization of\nthe Fr\\'echet distance (i.e., a morphing) that is locally optimal (aka locally\ncorrect), that is both easy to compute, and in practice, takes near linear time\non many inputs. The new morphing has the property that the leash is always as\nshort-as-possible.\n We implemented the new algorithm, and developed various strategies to get a\nfast execution in practice. Among our new contributions is a new simplification\nstrategy that is distance-sensitive, and enables us to compute the exact\ncontinuous Fr\\'echet distance in near linear time in practice. We preformed\nextensive experiments on our new algorithm, and released \\texttt{Julia} and\n\\texttt{Python} packages with these new implementations.\n", " In this study, a hybrid method based on an Adaptive Neuro-Fuzzy Inference\nSystem (ANFIS) and Particle Swarm Optimization (PSO) for diagnosing Liver\ndisorders (ANFIS-PSO) is introduced. This smart diagnosis method deals with a\ncombination of making an inference system and optimization process which tries\nto tune the hyper-parameters of ANFIS based on the data-set. The Liver diseases\ncharacteristics are taken from the UCI Repository of Machine Learning\nDatabases. The number of these characteristic attributes are 7, and the sample\nnumber is 354. The right diagnosis performance of the ANFIS-PSO intelligent\nmedical system for liver disease is evaluated by using classification accuracy,\nsensitivity and specificity analysis, respectively. According to the\nexperimental results, the performance of ANFIS-PSO can be more considerable\nthan traditional FIS and ANFIS without optimization phase.\n", " Markov Chain Monte Carlo (MCMC) techniques have long been studied in\ncomputational geometry subjects whereabouts the problems to be studied are\ncomplex geometric objects which by their nature require optimized techniques to\nbe deployed or to gain useful insights by them. MCMC approaches are directly\nanswering to geometric problems we are attempting to answer, and how these\nproblems could be deployed from theory to practice. Polytope which is a limited\nvolume in n-dimensional space specified by a collection of linear inequality\nconstraints require specific approximation. Therefore, sampling across density\nbased polytopes can not be performed without the use of such methods in which\nthe amount of repetition required is defined as a property of error margin. In\nthis work we propose a simple accurate sampling approach based on the\ntriangulation (tessellation) of a polytope. Moreover, we propose an efficient\nalgorithm named Density Based Sampling on Polytopes (DBSOP) for speedy MCMC\nsampling where the time required to perform sampling is significantly lower\ncompared to existing approaches in low dimensions with complexity\n$\\mathcal{O}^{*}\\left(n^{3}\\right)$. Ultimately, we highlight possible future\naspects and how the proposed scheme can be further improved with the\nintegration of reservoir-sampling based methods resulting in more speedy and\nefficient solution.\n" ]

[ " The aim of this work is to give degree formulas for the generalized Hamming\nweights of evaluation codes and to show lower bounds for these weights. In\nparticular, we give degree formulas for the generalized Hamming weights of\nReed--Muller-type codes, and we determine the minimum distance of toric codes\nover hypersimplices, and the 1st and 2nd generalized Hamming weights of\nsquarefree evaluation codes.\n", " The exploration of linear subspaces, particularly scattered subspaces, has\ngarnered considerable attention across diverse mathematical disciplines in\nrecent years, notably within finite geometries and coding theory. Scattered\nsubspaces play a pivotal role in analyzing various geometric structures such as\nblocking sets, two-intersection sets, complete arcs, caps in affine and\nprojective spaces over finite fields and rank metric codes. This paper\nintroduces a new infinite family of $h$-subspaces, along with their associated\nMRD codes. Additionally, it addresses the task of determining the generalized\nweights of these codes. Notably, we demonstrate that these MRD codes exhibit\nsome larger generalized weights compared to those previously identified.\n", " Simple drawings are drawings of graphs in the plane or on the sphere such\nthat vertices are distinct points, edges are Jordan arcs connecting their\nendpoints, and edges intersect at most once (either in a proper crossing or in\na shared endpoint). Simple drawings are generalized twisted if there is a point\n$O$ such that every ray emanating from $O$ crosses every edge of the drawing at\nmost once and there is a ray emanating from $O$ which crosses every edge\nexactly once. We show that all generalized twisted drawings of $K_n$ contain\nexactly $2n-4$ empty triangles, by this making a substantial step towards\nproving the conjecture that this is the case for every simple drawing of $K_n$.\n", " A new generalization of the Gray map is introduced. The new generalization\n$\\Phi: Z_{2^k}^n \\to Z_{2}^{2^{k-1}n}$ is connected with the known generalized\nGray map $\\phi$ in the following way: if we take two dual linear\n$Z_{2^k}$-codes and construct binary codes from them using the generalizations\n$\\phi$ and $\\Phi$ of the Gray map, then the weight enumerators of the binary\ncodes obtained will satisfy the MacWilliams identity. The classes of\n$Z_{2^k}$-linear Hadamard codes and co-$Z_{2^k}$-linear extended 1-perfect\ncodes are described, where co-$Z_{2^k}$-linearity means that the code can be\nobtained from a linear $Z_{2^k}$-code with the help of the new generalized Gray\nmap. Keywords: Gray map, Hadamard codes, MacWilliams identity, perfect codes,\n$Z_{2^k}$-linearity\n", " In the present paper we introduce and study finite point subsets of a special\nkind, called optimum distributions, in the n-dimensional unit cube. Such\ndistributions are closely related with known (delta,s,n)-nets of low\ndiscrepancy. It turns out that optimum distributions have a rich combinatorial\nstructure. Namely, we show that optimum distributions can be characterized\ncompletely as maximum distance separable codes with respect to a non-Hamming\nmetric. Weight spectra of such codes can be evaluated precisely. We also\nconsider linear codes and distributions and study their general properties\nincluding the duality with respect to a suitable inner product. The\ncorresponding generalized MacWilliams identities for weight enumerators are\nbriefly discussed. Broad classes of linear maximum distance separable codes and\nlinear optimum distributions are explicitely constructed in the paper by the\nHermite interpolations over finite fields.\n", " The paper studies properties of twisted sums of a Banach space $X$ with\n$c_0(\\kappa)$. We first prove a representation theorem for such twisted sums\nfrom which we will obtain, among others, the following: (a) twisted sums of\n$c_0(I)$ and $c_0(\\kappa)$ are either subspaces of $\\ell_\\infty(\\kappa)$ or\ntrivial on a copy of $c_0(\\kappa^+)$; (b) under the hypothesis $[\\mathfrak p =\n\\mathfrak c]$, when $K$ is either a suitable Corson compact, a separable\nRosenthal compact or a scattered compact of finite height, there is a twisted\nsum of $C(K)$ with $c_0(\\kappa)$ that is not isomorphic to a space of\ncontinuous functions; (c) all such twisted sums are Lindenstrauss spaces when\n$X$ is a Lindenstrauss space and $G$-spaces when $X=C(K)$ with $K$ convex,\nwhich shows tat a result of Benyamini is optimal; (d) they are isomorphically\npolyhedral when $X$ is a polyhedral space with property ($\\star$), which solves\na problem of Castillo and Papini.\n", " Convolutional codes are constructed, designed and analysed using row and/or\nblock structures of unit algebraic schemes. Infinite series of such codes and\nof codes with specific properties are derived. Properties are shown\nalgebraically and algebraic decoding methods are derived. For a given rate and\ngiven error-correction capability at each component, convolutional codes with\nthese specifications and with efficient decoding algorithms are constructed.\nExplicit prototype examples are given but in general large lengths and large\nerror capability are achievable. Convolutional codes with efficient decoding\nalgorithms at or near the maximum free distances attainable for the parameters\nare constructible. Unit memory convolutional codes of maximum possible free\ndistance are designed with practical algebraic decoding algorithms.\n LDPC (low density parity check) convolutional codes with efficient decoding\nschemes are constructed and analysed by the methods. Self-dual and\ndual-containing convolutional codes may also be designed by the methods;\ndual-containing codes enables the construction of quantum codes.\n", " In this communication, we present the results of the triple differential\ncross-section (TDCS) for the (e,2e) process on H2O molecule for the plane wave\nand the twisted electron beam impact. The formalism is developed in the first\nBorn approximation. We describe the plane/twisted wave, plane wave, the linear\ncombination of atomic orbitals (LCAO) (self-consistent field LCAO method) and\nCoulomb wave for the incident electron, scattered electron, the molecular state\nof H2O, and the ejected electron, respectively. We investigate the angular\nprofiles of the TDCS for the outer orbitals 1B1, 3A1, 1B2, and 2A1 of the water\nmolecule. We compare the angular profiles of the TDCS for the different values\nof orbital angular momentum (OAM) number m of the twisted electron beam with\nthat of the plane wave beam. We also study the TDCS for macroscopic H2O target\nto explore the effect of opening angle {\\theta}p of the twisted electron beam\non the TDCS. Our results clearly show the effect of the twisted electron's OAM\nnumber (m) and the opening angle {\\theta}p on the TDCS of the water molecule.\n", " In this article we present a class of codes with few weights arising from\nspecial type of linear sets. We explicitly show the weights of such codes,\ntheir weight enumerator and possible choices for their generator matrices. In\nparticular, our construction yields also to linear codes with three weights\nand, in some cases, to almost MDS codes. The interest for these codes relies on\ntheir applications to authentication codes and secret schemes, and their\nconnections with further objects such as association schemes and graphs.\n", " We study the spectrum of the Dirichlet Laplacian operator in a\ntwo-dimensional twisted strip embedded in $\\mathbb R^d$ with $d \\geq 2$. It is\nshown that a local twisting perturbation can create discrete eigenvalues for\nthe operator. In particular, we also study the case where the twisted effect\n\"grows\" at infinity while the width of the strip goes to zero. In this\nsituation, we find an asymptotic behavior for the eigenvalues.\n", " We introduce operator-valued twisted Araki-Woods algebras. These are\noperator-valued versions of a class of second quantization algebras that\nincludes $q$-Gaussian and $q$-Araki-Woods algebras and also generalize\nShlyakhtenko's von Neumann algebras generated by operator-valued semicircular\nvariables. We develop a disintegration theory that reduces the isomorphism type\nof operator-valued twisted Araki-Woods algebras over type I factors to the\nscalar-valued case. Moreover, these algebras come with a natural weight, and we\ncharacterize its modular theory. We also give sufficient criteria that\nguarantee that factoriality of these algebras.\n", " Huffman Codes are optimal Instantaneous Fixed-to-Variable (FV) codes in which\nevery source symbol can only be encoded by one codeword. Relaxing these\nconstraints permits constructing better FV codes. More specifically, recent\nwork has shown that AIFV-$m$ codes can beat Huffman coding. AIFV-$m$ codes\nconstruct am $m$-tuple of different coding trees between which the code\nalternates and are only almost instantaneous (AI). This means that decoding a\nword might require a delay of a finite number of bits.\n Current algorithms for constructing optimal AIFV-$m$ codes are iterative\nprocesses that construct progressively \"better sets\" of code trees. The\nprocesses have been proven to finitely converge to the optimal code but with no\nknown bounds on the convergence rate.\n This paper derives a geometric interpretation of the space of binary AIFV-$2$\ncodes, permitting the development of the first polynomially time-bounded\niterative procedures for constructing optimal AIFV codes.\n We first show that a simple binary search procedure can replace the current\niterative process to construct optimal binary AIFV-$2$ codes. We then describe\nhow to frame the problem as a linear programming with an exponential number of\nconstraints but a polynomial-time separability oracle. This permits using the\nGr\\\"otschel, Lov\\'asz and Schrijver ellipsoid method to solve the problem in a\npolynomial number of steps. While more complicated, this second method has the\npotential to lead to a polynomial time algorithm to construct optimal AIFV-$m$\ncodes for general $m$.\n", " We consider the system $\\mathcal{F}_4(a,b,c)$ of differential equations\nannihilating Appell's hypergeometric series $F_4(a,b,c;x)$. We find the\nintegral representations for four linearly independent solutions expressed by\nthe hypergeometric series $F_4$. By using the intersection forms of twisted\n(co)homology groups associated with them, we provide the monodromy\nrepresentation of $\\mathcal{F}_4(a,b,c)$ and the twisted period relations for\nthe fundamental systems of solutions of $\\mathcal{F}_4$.\n", " In this paper, we define Abelian and consta-Abelian polyadic codes over rings\ndefined as affine algebras over chain rings. For that aim, we use the classical\nconstruction via splittings and multipliers of the underlying Abelian group. We\nalso derive some results on the structure of the associated polyadic codes and\nthe number of codes under these conditions.\n", " We report some computational results comparing parallel and sequential codes\nfor vertex/facet enumeration problems for convex polyhedra. The problems chosen\nspan the range from simple to highly degenerate polytopes. We tested one code\n(lrs) based on pivoting and four codes (cddr+, ppl, normaliz, PORTA) based on\nthe double description method. normaliz employs parallelization as do the codes\nplrs and mplrs which are based on lrs. We tested these codes using various\nhardware configurations with up to 1200 cores. Major speedups were obtained by\nparallelization, particularly by the code mplrs which uses MPI and can operate\non clusters of machines.\n", " In this paper we study asymptotic distributions associated to piecewise\nquasi-polynomials. The main result obtained here is used in another paper of\nthe authors \"The equivariant index of twisted Dirac operators and\nsemi-classical limits\".\n", " Neural codes serve as a language for neurons in the brain. Convex codes,\nwhich arise from the pattern of intersections of convex sets in Euclidean\nspace, are of particular relevance to neuroscience. Not every code is convex,\nhowever, and the combinatorial properties of a code that determine its\nconvexity are still poorly understood. Here we find that a code that can be\nrealized by a collection of open convex sets may or may not be realizable by\nclosed convex sets, and vice versa, establishing that open convex and closed\nconvex codes are distinct classes. We also prove that max intersection-complete\ncodes (i.e. codes that contain all intersections of maximal codewords) are both\nopen convex and closed convex, and provide an upper bound for their minimal\nembedding dimension. Finally, we show that the addition of non-maximal\ncodewords to an open convex code preserves convexity.\n", " In this paper we analyse the construction of identification codes.\nIdentification codes are based on the question \"Is the message I have just\nreceived the one I am interested in?\", as opposed to Shannon's transmission,\nwhere the receiver is interested in not only one, but any message. The\nadvantage of identification is that it allows rates growing double\nexponentially in the blocklength at the cost of not being able to decode every\nmessage, which might be beneficial in certain applications. We focus on a\nspecial identification code construction based on two concatenated Reed-Solomon\ncodes and have a closer look at its implementation, analyzing the trade-offs of\nidentification with respect to transmission and the trade-offs introduced by\nthe computational cost of identification codes.\n", " We give sufficient conditions on the Lebesgue exponents for compositions of\nodd numbers of pseudo-differential operators with symbols in modulation spaces.\nAs a byproduct, we obtain sufficient conditions for twisted convolutions of odd\nnumbers of factors to be bounded on Wiener amalgam spaces.\n", " In this paper we give a partial shift version of user-irrepressible sequence\nsets and conflict-avoiding codes. By means of disjoint difference sets, we\nobtain an infinite number of such user-irrepressible sequence sets whose\nlengths are shorter than known results in general. Subsequently, the newly\ndefined partially conflict-avoiding codes are discussed.\n", " Using the concept of projective systems for linear codes and elementary\nlinear algebra, we show that projective $[n,k]_q$ codes form a connected\nsubgraph in the Grassmann graph consisting of $k$-dimensional subspaces of an\n$n$-dimensional vector space over the $q$-element field.\n", " An alternative proof is given of the existence of greatest lower bounds in\nthe imbalance order of binary maximal instantaneous codes of a given size.\nThese codes are viewed as maximal antichains of a given size in the infinite\nbinary tree of 0-1 words. The proof proposed makes use of a single balancing\noperation instead of expansion and contraction as in the original proof of the\nexistence of glb.\n", " Property testing has been a major area of research in computer science in the\nlast three decades. By property testing we refer to an ensemble of problems,\nresults and algorithms which enable to deduce global information about some\ndata by only reading small random parts of it. In recent years, this theory\nfound its way into group theory, mainly via group stability.\n In this paper, we study the following problem: Devise a randomized algorithm\nthat given a subgroup $H$ of $G$, decides whether $H$ is the whole group or a\nproper subgroup, by checking whether a single (random) element of $G$ is in\n$H$. The search for such an algorithm boils down to the following purely group\ntheoretic problem: For $G$ of rank $k$, find a small as possible test subset\n$A\\subseteq G$ such that for every proper subgroup $H$, $|H\\cap A|\\leq\n(1-\\delta)|A|$ for some absolute constant $\\delta>0$, which we call the\ndetection probability of $A$. It turns out that the search for sets $A$ of size\nlinear in $k$ and constant detection probability is a non-commutative analogue\nof the classical search for families of good error correcting codes. This paper\nis devoted to proving that such test subsets exist, which implies good\nuniversal error correcting codes exist -- providing a far reaching\ngeneralization of the classical result of Shannon.\n In addition, we study this problem in certain subclasses of groups -- such as\nabelian, nilpotent, and finite solvable groups -- providing different\nconstructions of test subsets for these subclasses with various qualities.\nFinally, this generalized theory of non-commutative error correcting codes\nsuggests a plethora of interesting problems and research directions.\n", " In this paper we characterize the algebraic structure of two-dimensional\n$(\\alpha,\\beta )$-constacyclic codes of arbitrary length $s.\\ell$ and of their\nduals. For $\\alpha,\\beta \\in \\{1,-1\\}$, we give necessary and sufficient\nconditions for a two-dimensional $(\\alpha,\\beta )$-constacyclic code to be\nself-dual. We also show that a two-dimensional $(\\alpha,1 )$-constacyclic code\n$\\mathcal{C}$ of length $n=s.\\ell$ can not be self-dual if $\\gcd(s,q)= 1$.\nFinally, we give some examples of self-dual, isodual, MDS and quasi-twisted\ncodes corresponding to two-dimensional $(\\alpha,\\beta )$-constacyclic codes.\n", " We discuss, in terms of rate-distortion theory, the fitness of molecular\ncodes as the problem of designing an optimal information channel. The fitness\nis governed by an interplay between the cost and quality of the channel, which\ninduces smoothness in the code. By incorporating this code fitness into\npopulation dynamics models, we suggest that the emergence and evolution of\nmolecular codes may be explained by simple channel design considerations.\n", " In this paper we show that under suitable conditions on their Fourier--Bohr\ncoefficients, the twisted Eberlein convolution of a measure with pure point\ndiffraction spectra and a measure with continuous diffraction spectra is zero.\nIn particular, the diffraction spectrum of a linear combinations of the two\nmeasures is simply the linear combinations of the two diffraction spectra with\nabsolute value square coefficients.\n", " A computational strategy and a collection of codes are presented for studying\nmultiparticle final states in hard hadronic collisions.\n", " The status of the theoretical uncertainties for LEP1 observables associated\nwith the corresponding comparison among different codes is briefly reviewed.\n", " Recent work of Cesana, Craig and the third author shows that the trace of\nplane partitions is asymptotically equidistributed in residue classes mod $b$.\nApplying a technique of the first two authors and Garnowski, we prove\nasymptotic formulas for the secondary terms in this equidistribution, which are\ncontrolled by certain complex numbers generated by a twisted MacMahon-type\nproduct. We further carry out a similar analysis for a statistic related to\nplane overpartitions.\n", " In this paper we study the main characteristics of some evaluation codes\nparameterized by the edges of a bipartite graph with a perfect matching.\n", " On a Goppa code whose structure polynomial has coefficients in the symbol\nfield, the Frobenius acts. Its fixed codewords form a subcode. Deleting the\nnaturally occurred redundance, we obtain a new code. It is proved that these\nnew codes approach the Gilbert-Varshamov bound. It is also proved that these\ncodes can be decoded within $O(n^2(\\logn)^a)$ operations in the symbol field,\nwhich is usually much small than the location field, where $n$ is the codeword\nlength, and $a$ a constant determined by the polynomial factorization\nalgorithm.\n", " Previous work on convexity of neural codes has produced codes that are\nopen-convex but not closed-convex -- or vice-versa. However, why a code is one\nbut not the other, and how to detect such discrepancies are open questions. We\ntackle these questions in two ways. First, we investigate the concept of\ndegeneracy introduced by Cruz et al., and extend their results to show that\nnondegeneracy precisely captures the situation when taking closures or\ninteriors of open or closed realizations, respectively, yields another\nrealization of the code. Second, we give the first general criteria for\nprecluding a code from being closed-convex (without ruling out open-convexity),\nunifying ad-hoc geometric arguments in prior works. One criterion is built on a\nphenomenon we call a rigid structure, while the other can be stated\nalgebraically, in terms of the neural ideal of the code. These results\ncomplement existing criteria having the opposite purpose: precluding\nopen-convexity but not closed-convexity. Finally, we show that a family of\ncodes shown by Jeffs to be not open-convex is in fact closed-convex and\nrealizable in dimension two.\n", " In order to facilitate the production of codes using WiSARD-based models,\nLabZero developed an ML library C++/Python called wisardpkg. This library is an\nMIT-licensed open-source package hosted on GitHub under the license.\n", " Over the past few decades, there has been extensive research on scattered\nsubspaces, partly because of their link to MRD codes. These subspaces can be\ncharacterized using linearized polynomials over finite fields. Within this\ncontext, scattered sequences extend the concept of scattered polynomials and\ncan be viewed as geometric equivalents of exceptional MRD codes. Up to now,\nonly scattered sequences of orders one and two have been developed. However,\nthis paper presents an infinite series of exceptional scattered sequences of\nany order beyond two which correspond to scattered subspaces that cannot be\nobtained as direct sum of scattered subspaces in smaller dimensions. The paper\nalso addresses equivalence concerns within this framework.\n", " In this work, we present a new version of non-binary VT codes that are\ncapable of correcting a single deletion or single insertion. Moreover, we\nprovide the first known linear time algorithms that encode user messages into\nthese codes of length n over the $q$-ary alphabet for $q\\ge 2$ with at most\n$\\ceil{\\log_q n} + 1$ redundant symbols, while the optimal redundancy required\nis at least $\\log_q n + \\log_q (q - 1)$ symbols. Our designed encoder reduces\nthe redundancy of the best-known encoder of Tenengolts (1984) by at least\n$2+\\log_q(3)$ redundant symbols, or equivalently $2\\log_2 q+3$ redundant bits.\n", " A set $C$ of unit vectors in $\\mathbb{R}^d$ is called an $L$-spherical code\nif $x \\cdot y \\in L$ for any distinct $x,y$ in $C$. Spherical codes have been\nextensively studied since their introduction in the 1970's by Delsarte,\nGoethals and Seidel. In this note we prove a conjecture of Bukh on the maximum\nsize of spherical codes. In particular, we show that for any set of $k$ fixed\nangles, one can choose at most $O(d^k)$ lines in $\\mathbb{R}^d$ such that any\npair of them forms one of these angles.\n", " To address growth challenges facing large Data Centers and supercomputing\nclusters a new construction is presented for scalable, high throughput, low\nlatency networks. The resulting networks require 1.5-5 times fewer switches,\n2-6 times fewer cables, have 1.2-2 times lower latency and correspondingly\nlower congestion and packet losses than the best present or proposed networks\nproviding the same number of ports at the same total bisection. These advantage\nratios increase with network size. The key new ingredient is the exact\nequivalence discovered between the problem of maximizing network bisection for\nlarge classes of practically interesting Cayley graphs and the problem of\nmaximizing codeword distance for linear error correcting codes. Resulting\ntranslation recipe converts existent optimal error correcting codes into\noptimal throughput networks.\n", " A new approach for the approximation of the channel log-likelihood ratio\n(LLR) for wireless channels based on Taylor series is proposed. The\napproximation is applied to the uncorrelated flat Rayleigh fading channel with\nunknown channel state information at the receiver. It is shown that the\nproposed approximation greatly simplifies the calculation of channel LLRs, and\nyet provides results almost identical to those based on the exact calculation\nof channel LLRs. The results are obtained in the context of bit-interleaved\ncoded modulation (BICM) schemes with low-density parity-check (LDPC) codes, and\ninclude threshold calculations and error rate performance of finite-length\ncodes. Compared to the existing approximations, the proposed method is either\nsignificantly less complex, or considerably more accurate.\n", " Let $n$ be a positive integer and $I$ a $k$-subset of integers in $[0,n-1]$.\nGiven a $k$-tuple $A=(\\alpha_0, \\cdots, \\alpha_{k-1})\\in \\mathbb{F}^k_{q^n}$,\nlet $M_{A,I}$ denote the matrix $(\\alpha_i^{q^j})$ with $0\\leq i\\leq k-1$ and\n$j\\in I$. When $I=\\{0,1,\\cdots, k-1\\}$, $M_{A,I}$ is called a Moore matrix\nwhich was introduced by E. H. Moore in 1896. It is well known that the\ndeterminant of a Moore matrix equals $0$ if and only if $\\alpha_0,\\cdots,\n\\alpha_{k-1}$ are $\\mathbb{F}_q$-linearly dependent. We call $I$ that satisfies\nthis property a Moore exponent set. In fact, Moore exponent sets are equivalent\nto maximum rank-distance (MRD) code with maximum left and right idealisers over\nfinite fields. It is already known that $I=\\{0,\\cdots, k-1\\}$ is not the unique\nMoore exponent set, for instance, (generalized) Delsarte-Gabidulin codes and\nthe MRD codes recently discovered by Csajb\\'ok, Marino, Polverino and the\nsecond author both give rise to new Moore exponent sets. By using algebraic\ngeometry approach, we obtain an asymptotic classification result: for $q>5$, if\n$I$ is not an arithmetic progression, then there exist an integer $N$ depending\non $I$ such that $I$ is not a Moore exponent set provided that $n>N$.\n", " We construct bijections giving three \"codes\" for trees. These codes follow\nnaturally from the Matrix Tree Theorem of Tutte and have many advantages over\nthe one produced by Prufer in 1918. One algorithm gives explicitly a bijection\nthat is implicit in Orlin's manipulatorial proof of Cayley's formula (the\nformula was actually found first by Borchardt). Another is based on a proof of\nKnuth. The third is an implementation of Joyal's pseudo-bijective proof of the\nformula, and is equivalent to one previously found by Egecioglu and Remmel. In\neach case, we have at least two algorithms, one of which involves hands-on\nmanipulations of the tree while the other involves a combinatorial and linear\nalgebraic manipulation of a matrix.\n", " We prove that on the twisted cotangent bundle of a closed manifold with an\nexact magnetic form, a Hamiltonian system of a time-dependent Tonelli\nHamiltonian function possesses infinitely many brake orbits. More precisely, by\napplying Legendre transform we show that there are infinitely many symmetric\norbits of the dual Euler-Lagrange system on the configuration space. This\nresult contains an assertion for the existence of infinitely many symmetric\norbits of Tonelli Euler-Lagrange systems given by G. Lu at the end of [Lu09a,\nRemark 6.1]. In this paper, we will present a complete proof of this assertion.\n", " We describe a strategy for constructing codes for DNA-based information\nstorage by serial composition of weighted finite-state transducers. The\nresulting state machines can integrate correction of substitution errors;\nsynchronization by interleaving watermark and periodic marker signals;\nconversion from binary to ternary, quaternary or mixed-radix sequences via an\nefficient block code; encoding into a DNA sequence that avoids homopolymer,\ndinucleotide, or trinucleotide runs and other short local repeats; and\ndetection/correction of errors (including local duplications, burst deletions,\nand substitutions) that are characteristic of DNA sequencing technologies. We\npresent software implementing these codes, available at\ngithub.com/ihh/dnastore, with simulation results demonstrating that the\ngenerated DNA is free of short repeats and can be accurately decoded even in\nthe presence of substitutions, short duplications and deletions.\n", " In the present paper, we introduce the concept of harmonic Tutte polynomials\nof matroids and discuss some of their properties. In particular, we generalize\nGreene's theorem, thereby expressing harmonic weight enumerators of codes as\nevaluations of harmonic Tutte polynomials.\n", " In this paper, we introduce the concept of the complete cycle index and\ndiscuss a relation with the complete weight enumerator in coding theory. This\nwork was motivated by Cameron's lecture note \"Polynomial aspects of codes,\nmatroids and permutation groups.\"\n", " In our recent paper we characterized the extension property for symmetrized\nweight composition for linear codes over a module alphabet. Several\nimprovements for the case of vector space alphabets are given in this paper. A\ndetailed description of the property of $G$-pseudo-injectivity for vector\nspaces is made.\n", " In this paper, the generalized fractional integral operators of two\ngeneralized Mittag-Leffler type functions are investigated. The special cases\nof interest involve the generalized Fox--Wright function and the generalized\nM-series and K-function. In the next Section 2 we first recall some generalized\nfractional integral operators among the most widely used in fractional\ncalculus. Section 3 is devoted to the definitions of M-series and K-function\nand their relations to special functions. In Sections 4 and 5, effective\nfractional calculus of the generalized M-series and the K-function is carried\nout. The last section briefly concludes and opens up new perspectives. The\nresults established herein generalize recent properties of generalized\nMittag-Leffler type functions using left-and right-sided generalized fractional\ndifferintegral operators. The note results also in important applications in\nphysics and mathematical engineering.\n", " Starting with a supersymmetric U(N)xU(N) gauge theory built in N=1\nsuperspace, a nonsupersymmetric theory is obtained by ``twisting'' the gauginos\ninto a different representation of the group than the gauge bosons. Despite the\nfact that this twisting breaks supersymmetry, it is still possible to construct\nan action that is holomorphic and invariant to local ``twisted'' gauge\ntransformations in superspace. It is conjectured that these two properties may\nallow the theory to be free of quadratic divergences to all orders, despite a\nlack of supersymmetry. An explicit calculation shows that the theory is free of\nquadratic divergences to at least the two-loop order.\n", " In my masters thesis I prove a square root bound on the distance of\nhomological codes that come from two dimensional surfaces, as a result of the\nsystolic inequality. I also give a detailed version of M.H. Freedman's proof\nthat due to systolic freedom, this bound does not hold in higher dimensions.\n", " The class of poset metrics is very large and contains some interesting\nfamilies of metrics. A family of metrics, based on posets which are formed from\ndisjoint chains which have the same size, is examined. A necessary and\nsufficient condition, for the existence of perfect single-error-correcting\ncodes for such poset metrics, is proved.\n", " Papadima and Suciu proved an inequality between the ranks of the cohomology\ngroups of the Aomoto complex with finite field coefficients and the twisted\ncohomology groups, and conjectured that they are actually equal for certain\ncases associated with the Milnor fiber of the arrangement. Recently, an\narrangement (the icosidodecahedral arrangement) with the following two peculiar\nproperties was found: (i) the strict version of Papadima-Suciu's inequality\nholds, and (ii) the first integral homology of the Milnor fiber has a\nnon-trivial $2$-torsion. In this paper, we investigate the relationship between\nthese two properties for double covering spaces. We prove that (i) and (ii) are\nactually equivalent.\n" ]

[ " Scientists use mathematical modelling to understand and predict the\nproperties of complex physical systems. In highly parameterised models there\noften exist relationships between parameters over which model predictions are\nidentical, or nearly so. These are known as structural or practical\nunidentifiabilities, respectively. They are hard to diagnose and make reliable\nparameter estimation from data impossible. They furthermore imply the existence\nof an underlying model simplification. We describe a scalable method for\ndetecting unidentifiabilities, and the functional relations defining them, for\ngeneric models. This allows for model simplification, and appreciation of which\nparameters (or functions thereof) cannot be estimated from data. Our algorithm\ncan identify features such as redundant mechanisms and fast timescale\nsubsystems, as well as the regimes in which such approximations are valid. We\nbase our algorithm on a novel quantification of regional parametric\nsensitivity: multiscale sloppiness. Traditionally, the link between parametric\nsensitivity and the conditioning of the parameter estimation problem is made\nlocally, through the Fisher Information Matrix. This is valid in the regime of\ninfinitesimal measurement uncertainty. We demonstrate the duality between\nmultiscale sloppiness and the geometry of confidence regions surrounding\nparameter estimates made where measurement uncertainty is non-negligible.\nFurther theoretical relationships are provided linking multiscale sloppiness to\nthe Likelihood-ratio test. From this, we show that a local sensitivity analysis\n(as typically done) is insufficient for determining the reliability of\nparameter estimation, even with simple (non)linear systems. Our algorithm\nprovides a tractable alternative. We finally apply our methods to a\nlarge-scale, benchmark Systems Biology model of NF-$\\kappa$B, uncovering\npreviously unknown unidentifiabilities.\n", " We consider high dimensional Wishart matrices $\\mathbb{X} \\mathbb{X}^{\\top}$\nwhere the entries of $\\mathbb{X} \\in {\\mathbb{R}^{n \\times d}}$ are i.i.d. from\na log-concave distribution. We prove an information theoretic phase transition:\nsuch matrices are close in total variation distance to the corresponding\nGaussian ensemble if and only if $d$ is much larger than $n^3$. Our proof is\nentropy-based, making use of the chain rule for relative entropy along with the\nrecursive structure in the definition of the Wishart ensemble. The proof\ncrucially relies on the well known relation between Fisher information and\nentropy, a variational representation for Fisher information, concentration\nbounds for the spectral norm of a random matrix, and certain small ball\nprobability estimates for log-concave measures.\n", " We point out an error in the proof of the main result of the paper of Tanabe\net al. (2007) concerning a parameter estimation for von Mises--Fisher\ndistributions, we correct the proof of the main result and we present a short\nalternative proof.\n", " We address metrological protocols for the estimation of the intensity and the\norientation of a magnetic field, and show that quantum-enhanced precision may\nbe achieved by probing the field with an arbitrary spin at thermal equilibrium.\nWe derive a general expression for the ultimate achievable precision, as given\nby the quantum Fisher information, and express this quantity in terms of common\nthermodynamic quantities. We also seek for the optimal observable, and show\nthat it corresponds to the spin projection along a suitable direction, defined\nby a universal function of the spin temperature. Finally, we prove the\nrobustness of our scheme against deviations of the measured spin projection\nfrom optimality.\n", " Normal estimation on 3D point clouds is a fundamental problem in 3D vision\nand graphics. Current methods often show limited accuracy in predicting normals\nat sharp features (e.g., edges and corners) and less robustness to noise. In\nthis paper, we propose a novel normal estimation method for point clouds. It\nconsists of two phases: (a) feature encoding which learns representations of\nlocal patches, and (b) normal estimation that takes the learned representation\nas input and regresses the normal vector. We are motivated that local patches\non isotropic and anisotropic surfaces have similar or distinct normals, and\nthat separable features or representations can be learned to facilitate normal\nestimation. To realise this, we first construct triplets of local patches on 3D\npoint cloud data, and design a triplet network with a triplet loss for feature\nencoding. We then design a simple network with several MLPs and a loss function\nto regress the normal vector. Despite having a smaller network size compared to\nmost other methods, experimental results show that our method preserves sharp\nfeatures and achieves better normal estimation results on CAD-like shapes.\n", " In this paper we study the presence of generalized blind spots, i.e. regions\nof the parameter space where the spin-independent cross section for dark matter\ndirect detection is suppressed, in the context of a generic 2HDM and a minimal\nfermionic Higgs-portal dark sector. To this end, we derive analytical\nexpressions for the couplings of the dark matter to the light and heavy\nHiggses, and thus for the blind spot solutions. Unlike the case of a standard\nHiggs sector, blind spots can occur even without a cancellation between\ndifferent contributions, while keeping unsuppressed and efficient the\nannihilation processes in the early Universe involving Higgs states. As a\nconsequence, the allowed parameter space is dramatically enhanced.\n", " Many applications, including rank aggregation, crowd-labeling, and graphon\nestimation, can be modeled in terms of a bivariate isotonic matrix with unknown\npermutations acting on its rows and/or columns. We consider the problem of\nestimating an unknown matrix in this class, based on noisy observations of\n(possibly, a subset of) its entries. We design and analyze polynomial-time\nalgorithms that improve upon the state of the art in two distinct metrics,\nshowing, in particular, that minimax optimal, computationally efficient\nestimation is achievable in certain settings. Along the way, we prove matching\nupper and lower bounds on the minimax radii of certain cone testing problems,\nwhich may be of independent interest.\n", " In this article, we explore the boundedness properties of pseudo-differential\noperators on radial sections of line bundles over the Poincar\\'e upper half\nplane, even when dealing with symbols of limited regularity. We first prove the\nboundedness of these operators when the symbol is smooth. To achieve this, we\nestablish a connection between the operator norm of the local part of our\npseudo-differential operators and the corresponding Euclidean\npseudo-differential operators. Additionally, we introduce a class of rough\nsymbols that lack any regularity conditions in the space variable and\ninvestigate the boundedness properties of the associated pseudo-differential\noperators. As a crucial part of our proof, we provide asymptotic estimates and\nfunctional identities for certain matrix coefficients of the principal and\ndiscrete series representations of the group $\\mathrm{SL(2,\\mathbb{R})}$.\n", " Large scale models of physical phenomena demand the development of new\nstatistical and computational tools in order to be effective. Many such models\nare `sloppy', i.e., exhibit behavior controlled by a relatively small number of\nparameter combinations. We review an information theoretic framework for\nanalyzing sloppy models. This formalism is based on the Fisher Information\nMatrix, which we interpret as a Riemannian metric on a parameterized space of\nmodels. Distance in this space is a measure of how distinguishable two models\nare based on their predictions. Sloppy model manifolds are bounded with a\nhierarchy of widths and extrinsic curvatures. We show how the manifold boundary\napproximation can extract the simple, hidden theory from complicated sloppy\nmodels. We attribute the success of simple effective models in physics as\nlikewise emerging from complicated processes exhibiting a low effective\ndimensionality. We discuss the ramifications and consequences of sloppy models\nfor biochemistry and science more generally. We suggest that the reason our\ncomplex world is understandable is due to the same fundamental reason: simple\ntheories of macroscopic behavior are hidden inside complicated microscopic\nprocesses.\n", " The formalisation of mathematics is continuing rapidly, however combinatorics\ncontinues to present challenges to formalisation efforts, such as its reliance\non techniques from a wide range of other fields in mathematics. This paper\npresents formal linear algebraic techniques for proofs on incidence structures\nin Isabelle/HOL, and their application to the first formalisation of Fisher's\ninequality. In addition to formalising incidence matrices and simple techniques\nfor reasoning on linear algebraic representations, the formalisation focuses on\nthe linear algebra bound and rank arguments. These techniques can easily be\nadapted for future formalisations in combinatorics, as we demonstrate through\nfurther application to proofs of variations on Fisher's inequality.\n", " In this paper we use Riesz spectral Theory and Gershgorin Theory to obtain\nexplicit information concerning the spectrum of pseudo-differential operators\ndefined on the unit circle $\\mathbb{T} := \\mathbb{R}/ 2 \\pi \\mathbb{ Z}$. For\nsymbols in the H\\\"ormander class $S^m_{1 , 0} (\\mathbb{T} \\times \\mathbb{Z})$,\nwe provide a sufficient and necessary condition to ensure that the\ncorresponding pseudo-differential operator is a Riesz operator in $L^p\n(\\mathbb{T})$, $1< p < \\infty$, extending in this way compact operators\ncharacterisation and Ghoberg's lemma to $L^p (\\mathbb{T})$. We provide an\nexample of a non-compact Riesz pseudo-differential operator in $L^p\n(\\mathbb{T})$, $1<p<2$. Also, for pseudo-differential operators with symbol\nsatisfying some integrability condition, it is defined its associated matrix in\nterms of the Fourier coefficients of the symbol, and this matrix is used to\ngive necessary and sufficient conditions for $L^2$-boundedness without assuming\nany regularity on the symbol, and to locate the spectrum of some operators.\n", " We consider several (related) notions of geometric regularity in the context\nof limit sets of geometrically finite Kleinian groups and associated\nPatterson-Sullivan measures. We begin by computing the upper and lower\nregularity dimensions of the Patterson-Sullivan measure, which involves\ncontrolling the relative measure of concentric balls. We then compute the\nAssouad and lower dimensions of the limit set, which involves controlling local\ndoubling properties. Unlike the Hausdorff, packing, and box-counting\ndimensions, we show that the Assouad and lower dimensions are not necessarily\ngiven by the Poincar\\'e exponent.\n", " Analog beamforming is an attractive and cost-effective solution to exploit\nthe benefits of massive multiple-input-multiple-output systems, by requiring\nonly one up/down-conversion chain. However, the presence of only one chain\nimposes a significant overhead in estimating the channel state information\nrequired for beamforming, when conventional digital channel estimation (CE)\napproaches are used. As an alternative, this paper proposes a novel CE\ntechnique, called periodic analog CE (PACE), that can be performed by analog\nhardware. By avoiding digital processing, the estimation overhead is\nsignificantly lowered and does not scale with number of antennas. PACE involves\nperiodic transmission of a sinusoidal reference signal by the transmitter,\nestimation of its amplitude and phase at each receive antenna via analog\nhardware, and using these estimates for beamforming. To enable such non-trivial\noperation, two reference tone recovery techniques and a novel receiver\narchitecture for PACE are proposed and analyzed, both theoretically and via\nsimulations. Results suggest that in sparse, wide-band channels and above a\ncertain signal-to-noise ratio, PACE aided beamforming suffers only a small loss\nin beamforming gain and enjoys a much lower CE overhead, in comparison to\nconventional approaches. Benefits of using PACE aided beamforming during the\ninitial access phase are also discussed.\n", " Over the years, scene understanding has attracted a growing interest in\ncomputer vision, providing the semantic and physical scene information\nnecessary for robots to complete some particular tasks autonomously. In 3D\nscenes, rich spatial geometric and topological information are often ignored by\nRGB-based approaches for scene understanding. In this study, we develop a\nbottom-up approach for scene understanding that infers support relations\nbetween objects from a point cloud. Our approach utilizes the spatial topology\ninformation of the plane pairs in the scene, consisting of three major steps.\n1) Detection of pairwise spatial configuration: dividing primitive pairs into\nlocal support connection and local inner connection; 2) primitive\nclassification: a combinatorial optimization method applied to classify\nprimitives; and 3) support relations inference and hierarchy graph\nconstruction: bottom-up support relations inference and scene hierarchy graph\nconstruction containing primitive level and object level. Through experiments,\nwe demonstrate that the algorithm achieves excellent performance in primitive\nclassification and support relations inference. Additionally, we show that the\nscene hierarchy graph contains rich geometric and topological information of\nobjects, and it possesses great scalability for scene understanding.\n", " In this paper, we study a stochastic disclosure control problem using\ninformation-theoretic methods. The useful data to be disclosed depend on\nprivate data that should be protected. Thus, we design a privacy mechanism to\nproduce new data which maximizes the disclosed information about the useful\ndata under a strong $\\chi^2$-privacy criterion. For sufficiently small leakage,\nthe privacy mechanism design problem can be geometrically studied in the space\nof probability distributions by a local approximation of the mutual\ninformation. By using methods from Euclidean information geometry, the original\nhighly challenging optimization problem can be reduced to a problem of finding\nthe principal right-singular vector of a matrix, which characterizes the\noptimal privacy mechanism. In two extensions we first consider a scenario where\nan adversary receives a noisy version of the user's message and then we look\nfor a mechanism which finds $U$ based on observing $X$, maximizing the mutual\ninformation between $U$ and $Y$ while satisfying the privacy criterion on $U$\nand $Z$ under the Markov chain $(Z,Y)-X-U$.\n", " We address the problem of restoring a high-quality image from an observed\nimage sequence strongly distorted by atmospheric turbulence. A novel algorithm\nis proposed in this paper to reduce geometric distortion as well as\nspace-and-time-varying blur due to strong turbulence. By considering a suitable\nenergy functional, our algorithm first obtains a sharp reference image and a\nsubsampled image sequence containing sharp and mildly distorted image frames\nwith respect to the reference image. The subsampled image sequence is then\nstabilized by applying the Robust Principal Component Analysis (RPCA) on the\ndeformation fields between image frames and warping the image frames by a\nquasiconformal map associated with the low-rank part of the deformation matrix.\nAfter image frames are registered to the reference image, the low-rank part of\nthem are deblurred via a blind deconvolution, and the deblurred frames are then\nfused with the enhanced sparse part. Experiments have been carried out on both\nsynthetic and real turbulence-distorted video. Results demonstrate that our\nmethod is effective in alleviating distortions and blur, restoring image\ndetails and enhancing visual quality.\n", " Due to the rapid spread of rumors on social media, rumor detection has become\nan extremely important challenge. Recently, numerous rumor detection models\nwhich utilize textual information and the propagation structure of events have\nbeen proposed. However, these methods overlook the importance of semantic\nevolvement information of event in propagation process, which is often\nchallenging to be truly learned in supervised training paradigms and\ntraditional rumor detection methods. To address this issue, we propose a novel\nsemantic evolvement enhanced Graph Autoencoder for Rumor Detection (GARD) model\nin this paper. The model learns semantic evolvement information of events by\ncapturing local semantic changes and global semantic evolvement information\nthrough specific graph autoencoder and reconstruction strategies. By combining\nsemantic evolvement information and propagation structure information, the\nmodel achieves a comprehensive understanding of event propagation and perform\naccurate and robust detection, while also detecting rumors earlier by capturing\nsemantic evolvement information in the early stages. Moreover, in order to\nenhance the model's ability to learn the distinct patterns of rumors and\nnon-rumors, we introduce a uniformity regularizer to further improve the\nmodel's performance. Experimental results on three public benchmark datasets\nconfirm the superiority of our GARD method over the state-of-the-art approaches\nin both overall performance and early rumor detection.\n", " Triangular mass matrices for neutrinos and their charged partners contain\nfull information on neutrino mixing in a most concise form. Although the scheme\nis general and model independent, triangular matrices are typical for reducible\nbut indecomposable representations of graded Lie algebras which, in turn, are\ncharacteristic for the standard model in noncommutative geometry. The mixing\nmatrix responsible for neutrino oscillations is worked out analytically for two\nand three lepton families. The example of two families fixes the mixing angle\nto just about what is required by the Mikheyev-Smirnov-Wolfenstein resonance\noscillation of solar neutrinos. In the case of three families we classify all\nphysically plausible choices for the neutrino mass matrix and derive\ninteresting bounds on some of the moduli of the mixing matrix.\n", " Partition of unities appear in many places in analysis. Typically they are\ngenerated by compactly supported functions with a certain regularity. In this\npaper we consider partition of unities obtained as integer-translates of entire\nfunctions restricted to finite intervals. We characterize the entire functions\nthat lead to a partition of unity in this way, and we provide characterizations\nof the \"cut-off\" entire functions, considered as functions of a real variable,\nto have desired regularity. In particular we obtain partition of unities\ngenerated by functions with small support and desired regularity. Applied to\nGabor analysis this leads to constructions of dual pairs of Gabor frames with\nlow redundancy, generated by trigonometric polynomials with small support and\ndesired regularity.\n", " This paper reviews the main estimation and prediction results derived in the\ncontext of functional time series, when Hilbert and Banach spaces are\nconsidered, specially, in the context of autoregressive processes of order one\n(ARH(1) and ARB(1) processes, for H and B being a Hilbert and Banach space,\nrespectively). Particularly, we pay attention to the estimation and prediction\nresults, and statistical tests, derived in both parametric and non-parametric\nframeworks. A comparative study between different ARH(1) prediction approaches\nis developed in the simulation study undertaken.\n", " We study the existence problem for a local implicit function determined by a\nsystem of nonlinear algebraic equations in the particular case when the\ndeterminant of its Jacobian matrix vanishes at the point under consideration.\nWe present a system of sufficient conditions that implies existence of a local\nimplicit function as well as another system of sufficient conditions that\nguarantees absence of a local implicit function. The results obtained are\napplied to proving new and classical results on flexibility and rigidity of\npolyhedra and frameworks.\n", " In this paper, realistic CPT-symmetric united at substance quark and neutrino\nmixing matrices are studied by the idea that they are shifted from respectively\nunit and bimaximal benchmark matrices by one new small universal empirical\ne-parameter. At leading finite e-approximation quark-lepton complementarity\n(QLC) is an exact regularity. Equal solar and atmospheric deviations from exact\nQLC constitute one of the main inferences at next to leading approximation.\nThat universal violation of exact QLC advances the status of the QLC idea; and\nthe more so, as that deviation from QLC is quantitatively estimated and used\nfor accurate calculations of lepton mixing angles. Quark and neutrino Dirac\nCP-violating phases are determined by the considered quadratic hierarchy\nparadigm in flavor phenomenology. Inferences are partly supported by quark\nheavy flavor unitarity triangle angles and testable at B-factory and LHC b\nexperiments. Estimated magnitude of the unitarity triangle gamma-angle concurs\nwith the quark CP-violating phase to within ~5x10-4. The final quark CKM mixing\nmatrix in terms of the e-parameter is in excellent quantitative agreement with\nworld data and suggests a fitting form of neutrino PMNS mixing matrix. The\ne-constant implications in flavor phenomenology are considered.\n", " We focus on the problem of manifold estimation: given a set of observations\nsampled close to some unknown submanifold $M$, one wants to recover information\nabout the geometry of $M$. Minimax estimators which have been proposed so far\nall depend crucially on the a priori knowledge of some parameters quantifying\nthe underlying distribution generating the sample (such as bounds on its\ndensity), whereas those quantities will be unknown in practice. Our\ncontribution to the matter is twofold: first, we introduce a one-parameter\nfamily of manifold estimators $(\\hat{M}_t)_{t\\geq 0}$ based on a localized\nversion of convex hulls, and show that for some choice of $t$, the\ncorresponding estimator is minimax on the class of models of $C^2$ manifolds\nintroduced in [Genovese et al., Manifold estimation and singular deconvolution\nunder Hausdorff loss]. Second, we propose a completely data-driven selection\nprocedure for the parameter $t$, leading to a minimax adaptive manifold\nestimator on this class of models. This selection procedure actually allows us\nto recover the Hausdorff distance between the set of observations and $M$, and\ncan therefore be used as a scale parameter in other settings, such as tangent\nspace estimation.\n", " Bayesian inference provides a rigorous methodology for estimation and\nuncertainty quantification of parameters in geophysical forward models.\nBadlands (basin and landscape dynamics model) is a landscape evolution model\nthat simulates topography development at various space and time scales.\nBadlands consists of a number of geophysical parameters that needs estimation\nwith appropriate uncertainty quantification; given the observed present-day\nground truth such as surface topography and the stratigraphy of sediment\ndeposition through time. The inference of unknown parameters is challenging due\nto the scarcity of data, sensitivity of the parameter setting and complexity of\nthe Badlands model. In this paper, we take a Bayesian approach to provide\ninference using Markov chain Monte Carlo sampling (MCMC). We present\n\\textit{Bayeslands}; a Bayesian framework for Badlands that fuses information\nobtained from complex forward models with observational data and prior\nknowledge. As a proof-of-concept, we consider a synthetic and real-world\ntopography with two parameters for Bayeslands inference, namely precipitation\nand erodibility. The results of the experiments show that Bayeslands yields a\npromising distribution of the parameters. Moreover, we demonstrate the\nchallenge in sampling irregular and multi-modal posterior distributions using a\nlikelihood surface that has a range of sub-optimal modes.\n", " Partially-observed Boolean dynamical systems (POBDS) are a general class of\nnonlinear models with application in estimation and control of Boolean\nprocesses based on noisy and incomplete measurements. The optimal minimum mean\nsquare error (MMSE) algorithms for POBDS state estimation, namely, the Boolean\nKalman filter (BKF) and Boolean Kalman smoother (BKS), are intractable in the\ncase of large systems, due to computational and memory requirements. To address\nthis, we propose approximate MMSE filtering and smoothing algorithms based on\nthe auxiliary particle filter (APF) method from sequential Monte-Carlo theory.\nThese algorithms are used jointly with maximum-likelihood (ML) methods for\nsimultaneous state and parameter estimation in POBDS models. In the presence of\ncontinuous parameters, ML estimation is performed using the\nexpectation-maximization (EM) algorithm; we develop for this purpose a special\nsmoother which reduces the computational complexity of the EM algorithm. The\nresulting particle-based adaptive filter is applied to a POBDS model of Boolean\ngene regulatory networks observed through noisy RNA-Seq time series data, and\nperformance is assessed through a series of numerical experiments using the\nwell-known cell cycle gene regulatory model.\n", " Global fits to moments of kinematic distributions measured in inclusive\nsemileptonic $B\\rightarrow X_c l \\nu_l$ enable the determination of the\nCabibbo-Kobayashi-Maskawa matrix element $|V_{cb}|$ together with\nnon-perturbative matrix elements of the heavy quark expansion. In current fits,\nonly two distinct kinematic distributions are employed and, as a consequence,\nhigher moments of these distributions need to be taken into account to extract\nthe relevant non-perturbative matrix elements. The moments of a given\ndistribution are highly correlated and experimental uncertainties increase for\nhigher moments. To address these issues, Turczyk suggested the inclusion of the\ncharged lepton forward-backward asymmetry $\\mathcal{A}_{FB}$ in global fits,\nsince it provides information on non-perturbative parameters beyond the\ncommonly used moments.\n It is possible to construct differences of partial moments of kinematic\ndistributions, which can provide additional information on the non-perturbative\nparameters beyond $\\mathcal{A}_{FB}$ and are studied in this work for the first\ntime. Further, experimental cuts on the four-momentum transfer square are\nstudied and are shown to preserve the shape of the angular distribution, in\ncontrast to commonly used cuts on the lepton energy. Finally, the impact of\nfinal-state radiation and experimental lepton identification requirements on\nmeasurements of $\\mathcal{A}_{FB}$ and differences of partial moments are\ndiscussed.\n", " We give effectivized Holder-logarithmic energy and regularity dependent\nstability estimates for the Gel'fand inverse boundary value problem in\ndimension $d=3$. This effectivization includes explicit dependance of the\nestimates on coefficient norms and related parameters. Our new estimates are\ngiven in $L^2$ and $L^\\infty$ norms for the coefficient difference and related\nstability efficiently increases with increasing energy and/or coefficient\ndifference regularity. Comparisons with preceeding results are given.\n", " We propose a data-driven method for recovering miss-ing parts of 3D shapes.\nOur method is based on a new deep learning architecture consisting of two\nsub-networks: a global structure inference network and a local geometry\nrefinement network. The global structure inference network incorporates a long\nshort-term memorized context fusion module (LSTM-CF) that infers the global\nstructure of the shape based on multi-view depth information provided as part\nof the input. It also includes a 3D fully convolutional (3DFCN) module that\nfurther enriches the global structure representation according to volumetric\ninformation in the input. Under the guidance of the global structure network,\nthe local geometry refinement network takes as input lo-cal 3D patches around\nmissing regions, and progressively produces a high-resolution, complete surface\nthrough a volumetric encoder-decoder architecture. Our method jointly trains\nthe global structure inference and local geometry refinement networks in an\nend-to-end manner. We perform qualitative and quantitative evaluations on six\nobject categories, demonstrating that our method outperforms existing\nstate-of-the-art work on shape completion.\n", " We study planar \"vertex\" models, which are probability measures on edge\nsubsets of a planar graph, satisfying certain constraints at each vertex,\nexamples including dimer model, and 1-2 model, which we will define. We express\nthe local statistics of a large class of vertex models on a finite hexagonal\nlattice as a linear combination of the local statistics of dimers on the\ncorresponding Fisher graph, with the help of a generalized holographic\nalgorithm. Using an $n\\times n$ torus to approximate the periodic infinite\ngraph, we give an explicit integral formula for the free energy and local\nstatistics for configurations of the vertex model on an infinite bi-periodic\ngraph. As an example, we simulate the 1-2 model by the technique of Glauber\ndynamics.\n", " We study sparse hypergraphs which satisfy a mild pseudorandomness condition\nknown as $L_p$ regularity. We prove appropriate regularity and counting lemmas,\nand we extend the relative removal lemma of Tao in this setting. This answers a\nquestion of Borgs, Chayes, Cohn and Zhao.\n", " We present a study for the generation of events from a physical process with\ndeep generative models. The simulation of physical processes requires not only\nthe production of physical events, but also to ensure these events occur with\nthe correct frequencies. We investigate the feasibility of learning the event\ngeneration and the frequency of occurrence with Generative Adversarial Networks\n(GANs) and Variational Autoencoders (VAEs) to produce events like Monte Carlo\ngenerators. We study three processes: a simple two-body decay, the processes\n$e^+e^-\\to Z \\to l^+l^-$ and $p p \\to t\\bar{t} $ including the decay of the top\nquarks and a simulation of the detector response. We find that the tested GAN\narchitectures and the standard VAE are not able to learn the distributions\nprecisely. By buffering density information of encoded Monte Carlo events given\nthe encoder of a VAE we are able to construct a prior for the sampling of new\nevents from the decoder that yields distributions that are in very good\nagreement with real Monte Carlo events and are generated several orders of\nmagnitude faster. Applications of this work include generic density estimation\nand sampling, targeted event generation via a principal component analysis of\nencoded ground truth data, anomaly detection and more efficient importance\nsampling, e.g. for the phase space integration of matrix elements in quantum\nfield theories.\n", " Bayesian neural learning feature a rigorous approach to estimation and\nuncertainty quantification via the posterior distribution of weights that\nrepresent knowledge of the neural network. This not only provides point\nestimates of optimal set of weights but also the ability to quantify\nuncertainty in decision making using the posterior distribution. Markov chain\nMonte Carlo (MCMC) techniques are typically used to obtain sample-based\nestimates of the posterior distribution. However, these techniques face\nchallenges in convergence and scalability, particularly in settings with large\ndatasets and network architectures. This paper address these challenges in two\nways. First, parallel tempering is used used to explore multiple modes of the\nposterior distribution and implemented in multi-core computing architecture.\nSecond, we make within-chain sampling schemes more efficient by using Langevin\ngradient information in forming Metropolis-Hastings proposal distributions. We\ndemonstrate the techniques using time series prediction and pattern\nclassification applications. The results show that the method not only improves\nthe computational time, but provides better prediction or decision making\ncapabilities when compared to related methods.\n", " We establish universality of local eigenvalue correlations in unitary random\nmatrix ensembles (1/Z_n) |\\det M|^{2\\alpha} e^{-n\\tr V(M)} dM near the origin\nof the spectrum. If V is even, and if the recurrence coefficients of the\northogonal polynomials associated with |x|^{2\\alpha} e^{-nV(x)} have a regular\nlimiting behavior, then it is known from work of Akemann et al., and Kanzieper\nand Freilikher that the local eigenvalue correlations have universal behavior\ndescribed in terms of Bessel functions. We extend this to a much wider class of\nconfining potentials V. Our approach is based on the steepest descent method of\nDeift and Zhou for the asymptotic analysis of Riemann-Hilbert problems. This\nmethod was used by Deift et al. to establish universality in the bulk of the\nspectrum. A main part of the present work is devoted to the analysis of a local\nRiemann-Hilbert problem near the origin.\n", " We consider chemical reaction networks modeled by a discrete state and\ncontinuous in time Markov process for the vector copy number of the species and\nprovide a novel particle filter method for state and parameter estimation based\non exact observation of some of the species in continuous time. The conditional\nprobability distribution of the unobserved states is shown to satisfy a system\nof differential equations with jumps. We provide a method of simulating a\nprocess that is a proxy for the vector copy number of the unobserved species\nalong with a weight. The resulting weighted Monte Carlo simulation is then used\nto compute the conditional probability distribution of the unobserved species.\nWe also show how our algorithm can be adapted for a Bayesian estimation of\nparameters and for the estimation of a past state value based on observations\nup to a future time.\n", " Recent interest in integrated sensing and communications has led to the\ndesign of novel signal processing techniques to recover information from an\noverlaid radar-communications signal. Here, we focus on a spectral coexistence\nscenario, wherein the channels and transmit signals of both radar and\ncommunications systems are unknown to the common receiver. In this dual-blind\ndeconvolution (DBD) problem, the receiver admits a multi-carrier wireless\ncommunications signal that is overlaid with the radar signal reflected off\nmultiple targets. The communications and radar channels are represented by\ncontinuous-valued range-times or delays corresponding to multiple transmission\npaths and targets, respectively. Prior works addressed recovery of unknown\nchannels and signals in this ill-posed DBD problem through atomic norm\nminimization but contingent on individual minimum separation conditions for\nradar and communications channels. In this paper, we provide an optimal joint\nseparation condition using extremal functions from the Beurling-Selberg\ninterpolation theory. Thereafter, we formulate DBD as a low-rank modified\nHankel matrix retrieval and solve it via nuclear norm minimization. We estimate\nthe unknown target and communications parameters from the recovered low-rank\nmatrix using multiple signal classification (MUSIC) method. We show that the\njoint separation condition also guarantees that the underlying Vandermonde\nmatrix for MUSIC is well-conditioned. Numerical experiments validate our\ntheoretical findings.\n", " Nonuniformities in the imaging characteristics of modern image sensors are a\nprimary factor in the push to develop a pixel-level generalization of the\nphoton transfer characterization method. In this paper, we seek to develop a\nbody of theoretical results leading toward a comprehensive approach for\ntackling the biggest obstacle in the way of this goal: a means of pixel-level\nconversion gain estimation. This is accomplished by developing an estimator for\nthe reciprocal-difference of normal variances and then using this to construct\na novel estimator of the conversion gain. The first two moments of this\nestimator are derived and used to construct exact and approximate confidence\nintervals for its absolute relative bias and absolute coefficient of variation,\nrespectively. A means of approximating and computing optimal sample sizes are\nalso discussed and used to demonstrate the process of pixel-level conversion\ngain estimation for a real image sensor.\n", " The tacnode Riemann-Hilbert problem is a 4 x 4 matrix valued RH problem that\nappears in the description of the local behavior of two touching groups of\nnon-intersecting Brownian motions. The same RH problem was also found by Duits\nand Geudens to describe a new critical regime in the two-matrix model.\n Delvaux gave integral representations for some of the entries of the 4 x 4\nmatrix. We complement this work by presenting integral representations for all\nof the entries. As a consequence we give an explicit formula for the\nDuits-Geudens critical kernel.\n", " We describe the variety of fixed points of a unipotent operator acting on the\nspace of matrices. We compute the determinant and the rank of a generic\n(symmetric, or anti-symmetric) matrix in the fixed variety, yielding\ninformation about the generic singular locus of the corresponding bilinear\nform.\n", " The rational covariance extension problem (RCEP) is an important problem in\nsystems and control occurring in such diverse fields as control, estimation,\nsystem identification, and signal and image processing, leading to many\nfundamental theoretical questions. In fact, this inverse problem is a key\ncomponent in many identification and signal processing techniques and plays a\nfundamental role in prediction, analysis, and modeling of systems and signals.\nIt is well-known that the RCEP can be reformulated as a (truncated)\ntrigonometric moment problem subject to a rationality condition. In this paper\nwe consider the more general multidimensional trigonometric moment problem with\na similar rationality constraint. This generalization creates many interesting\nnew mathematical questions and also provides new insights into the original\none-dimensional problem. A key concept in this approach is the complete smooth\nparametrization of all solutions, allowing solutions to be tuned to satisfy\nadditional design specifications without violating the complexity constraints.\nAs an illustration of the potential of this approach we apply our results to\nmultidimensional spectral estimation and image compression. This is just a\nfirst step in this direction, and we expect that more elaborate tuning\nstrategies will enhance our procedures in the future.\n", " In the context of increasing demands for long-term multi-energy load\nforecasting in real-world applications, this paper introduces Patchformer, a\nnovel model that integrates patch embedding with encoder-decoder\nTransformer-based architectures. To address the limitation in existing\nTransformer-based models, which struggle with intricate temporal patterns in\nlong-term forecasting, Patchformer employs patch embedding, which predicts\nmultivariate time-series data by separating it into multiple univariate data\nand segmenting each of them into multiple patches. This method effectively\nenhances the model's ability to capture local and global semantic dependencies.\nThe numerical analysis shows that the Patchformer obtains overall better\nprediction accuracy in both multivariate and univariate long-term forecasting\non the novel Multi-Energy dataset and other benchmark datasets. In addition,\nthe positive effect of the interdependence among energy-related products on the\nperformance of long-term time-series forecasting across Patchformer and other\ncompared models is discovered, and the superiority of the Patchformer against\nother models is also demonstrated, which presents a significant advancement in\nhandling the interdependence and complexities of long-term multi-energy\nforecasting. Lastly, Patchformer is illustrated as the only model that follows\nthe positive correlation between model performance and the length of the past\nsequence, which states its ability to capture long-range past local semantic\ninformation.\n", " The Darlington synthesis problem (in the scalar case) is the problem of\nembedding a given contractive analytic function to an inner $2\\times 2$ matrix\nfunction as the entry. A fundamental result of Arov--Douglas--Helton relates\nthis algebraic property to a pure analytic one known as a pseudocontinuation of\nbounded type. We suggest a local version of the Darlington synthesis problem\nand prove a local analog of the ADH theorem.\n", " By extending the notion of Arens regularity of bilinear mappings, we say that\na bounded trilinear map on Banach spaces id Aron--Berner regular when all its\nsix Aron-Berner extensions to the bidual spaces coincide. We give some results\non the Aron-Berner regularity of certain trilinear maps. We then focus on the\nbidual, $E^{**},$ of a Jordan banach triple system $(E,\\pi)$, and investigate\nthose conditions under which $E^{**}$ is itself a Jordan Banach triple system\nunder each of the Aron-Berner extensions of the triple product $\\pi.$ We also\ncompare these six triple products with those arising from certain ultrafilters\nbased on the ultrapower formulation of the principle of local reflexivity. In\nparticular, we examine the Aron--Berner triple products on the bidual of a\nJB$^*$-triple in relation with the so-called Dineen's theorem. Some\nilluminating examples are included and some questions are also left undecided.\n", " We study the regularity of the bilinear maximal operator when applied to\nSobolev functions, proving that it maps $W^{1,p}(\\mathbb{R}) \\times\nW^{1,q}(\\mathbb{R}) \\to W^{1,r}(\\mathbb{R})$ with $1 <p,q < \\infty$ and $r\\geq\n1$, boundedly and continuously. The same result holds on $\\mathbb{R}^n$ when\n$r>1$. We also investigate the almost everywhere and weak convergence under the\naction of the classical Hardy-Littlewood maximal operator, both in its global\nand local versions.\n", " Scientific data has been growing in both size and complexity across the\nmodern physical, engineering, life and social sciences. Spatial structure, for\nexample, is a hallmark of many of the most important real-world complex\nsystems, but its analysis is fraught with statistical challenges. Topological\ndata analysis can provide a powerful computational window on complex systems.\nHere we present a framework to extend and interpret persistent homology\nsummaries to analyse spatial data across multiple scales. We introduce\nhyperTDA, a topological pipeline that unifies local (e.g. geodesic) and global\n(e.g. Euclidean) metrics without losing spatial information, even in the\npresence of noise. Homology generators offer an elegant and flexible\ndescription of spatial structures and can capture the information computed by\npersistent homology in an interpretable way. Here the information computed by\npersistent homology is transformed into a weighted hypergraph, where hyperedges\ncorrespond to homology generators. We consider different choices of generators\n(e.g. matroid or minimal) and find that centrality and community detection are\nrobust to either choice. We compare hyperTDA to existing geometric measures and\nvalidate its robustness to noise. We demonstrate the power of computing\nhigher-order topological structures on spatial curves arising frequently in\necology, biophysics, and biology, but also in high-dimensional financial\ndatasets. We find that hyperTDA can select between synthetic trajectories from\nthe landmark 2020 AnDi challenge and quantifies movements of different animal\nspecies, even when data is limited.\n", " We consider tree-dimensional domino tilings of cylinders $\\mathcal{D} \\times\n[0,N] \\subset \\mathbb{R}^3$, where $\\mathcal{D} \\subset \\mathbb{R}^2$ is a\nbalanced quadriculated disk and $N \\in \\mathbb{N}$. A flip is a local move in\nthe space of tilings: two adjacent and parallel dominoes are removed and then\nreplaced in a different position. The twist is a flip invariant that associates\nan integer number to a domino tiling. A disk $\\mathcal{D}$ is called regular if\nany two tilings of $\\mathcal{D} \\times [0,N]$ sharing the same twist can be\nconnected trough a sequence of flips once extra vertical space is added to the\ncylinder. We prove that hamiltonian disks with narrow and small bottlenecks are\nregular. In particular, we show that the absence of a bottleneck in a\nhamiltonian disk implies regularity.\n", " A fundamental problem in drug discovery is to design molecules that bind to\nspecific proteins. To tackle this problem using machine learning methods, here\nwe propose a novel and effective framework, known as GraphBP, to generate 3D\nmolecules that bind to given proteins by placing atoms of specific types and\nlocations to the given binding site one by one. In particular, at each step, we\nfirst employ a 3D graph neural network to obtain geometry-aware and chemically\ninformative representations from the intermediate contextual information. Such\ncontext includes the given binding site and atoms placed in the previous steps.\nSecond, to preserve the desirable equivariance property, we select a local\nreference atom according to the designed auxiliary classifiers and then\nconstruct a local spherical coordinate system. Finally, to place a new atom, we\ngenerate its atom type and relative location w.r.t. the constructed local\ncoordinate system via a flow model. We also consider generating the variables\nof interest sequentially to capture the underlying dependencies among them.\nExperiments demonstrate that our GraphBP is effective to generate 3D molecules\nwith binding ability to target protein binding sites. Our implementation is\navailable at https://github.com/divelab/GraphBP.\n", " Families of ODE models characterized by a common sign structure of their\nJacobi matrix are investigated within the formalism of qualitative differential\nequations. In the context of regulatory networks the sign structure of the\nJacobi matrix carries the information about which components of the network\ninhibit or activate each other. Information about constraints on the behavior\nof models in this family is stored in a so called qualitative state transition\ngraph. We showed previously that a similar approach can be used to analyze a\nmodel pool of Boolean functions characterized by a common interaction graph.\nHere we show that the opposite approach is fruitful as well. We show that the\nqualitative state transition graph can be reduced to a \"skeleton\" represented\nby a Boolean function conserving the reachability properties. This reduction\nhas the advantage that approaches such as model checking and network inference\nmethods can be applied to the \"skeleton\" within the framework of Boolean\nnetworks. Furthermore, our work constitutes an alternative to approach to link\nBoolean networks and differential equations.\n", " An exact and general expression for the analytic wavelet transform of a\nreal-valued signal is constructed, resolving the time-dependent effects of\nnon-negligible amplitude and frequency modulation. The analytic signal is first\nlocally represented as a modulated oscillation, demodulated by its own\ninstantaneous frequency, and then Taylor-expanded at each point in time. The\nterms in this expansion, called the instantaneous modulation functions, are\ntime-varying functions which quantify, at increasingly higher orders, the local\ndepartures of the signal from a uniform sinusoidal oscillation. Closed-form\nexpressions for these functions are found in terms of Bell polynomials and\nderivatives of the signal's instantaneous frequency and bandwidth. The analytic\nwavelet transform is shown to depend upon the interaction between the signal's\ninstantaneous modulation functions and frequency-domain derivatives of the\nwavelet, inducing a hierarchy of departures of the transform away from a\nperfect representation of the signal. The form of these deviation terms\nsuggests a set of conditions for matching the wavelet properties to suit the\nvariability of the signal, in which case our expressions simplify considerably.\nOne may then quantify the time-varying bias associated with signal estimation\nvia wavelet ridge analysis, and choose wavelets to minimize this bias.\n", " In this paper, we study different types of weighted Besov and\nTriebel-Lizorkin spaces with variable smoothness. The function spaces can be\ndefined by means of the Littlewood-Paley theory in the field of Fourier\nanalysis, while there are other norms arising in the theory of partial\ndifferential equations such as Sobolev-Slobodeckij spaces. It is known that two\nnorms are equivalent when one considers constant regularity function spaces\nwithout weights. We show that the equivalence still holds for variable\nsmoothness and weights, which is accomplished by making use of shifted maximal\nfunctions, Peetre's maximal functions, and the reverse H\\\"older inequality.\nMoreover, we obtain a weighted regularity estimate for time-fractional\nevolution equations and a generalized Sobolev embedding theorem without\nweights.\n", " A matrix is given in ``shredded'' form if we are presented with the multiset\nof rows and the multiset of columns, but not told which row is which or which\ncolumn is which. The matrix is reconstructible if it is uniquely determined by\nthis information. Let $M$ be a random binary $n\\times n$ matrix, where each\nentry independently is $1$ with probability $p=p(n)\\le\\frac12$. Atamanchuk,\nDevroye and Vicenzo introduced the problem and showed that $M$ is\nreconstructible with high probability for $p\\ge (2+\\varepsilon)\\frac{1}{n}\\log\nn$. Here we find that the sharp threshold for reconstructibility is at\n$p\\sim\\frac{1}{2n}\\log n$.\n" ]

[ " The task of finding the optimal compression of a polyline with straight-line\nsegments and arcs is performed in many applications, such as polyline\ncompression, noise filtering, and feature recognition. Optimal compression\nalgorithms find the best solution using the dynamic programming approach, which\nrequires a significant amount of arc fitting. This paper describes an\nimprovement to the dynamic programming approach by reducing the amount of arc\nfitting necessary to find the optimal solution. Instead of processing from the\nsecond to the last vertices in the dynamic programming approach, the algorithm\nproceeds forward and skips as many steps as possible without affecting the\ninference in any way. Such a modification extends the practical application of\nthe algorithm to polylines having arcs with a large number of vertices.\n", " In the case when the weight and its inverse belong to BMO(T), we prove the\nasymptotics of the monic orthogonal polynomials in L^p, 2<p<p_0. Immediate\napplications include the estimates on the uniform norm and asymptotics for the\npolynomial entropy.\n", " A compression algorithm is presented that uses the set of prime numbers.\nSequences of numbers are correlated with the prime numbers, and labeled with\nthe integers. The algorithm can be iterated on data sets, generating factors of\ndoubles on the compression.\n", " We obtain asymptotic expansions of the spatially discrete 2D heat kernels, or\nGreen's functions on lattices, with respect to powers of time variable up to an\narbitrary order and estimate the remainders uniformly on the whole lattice.\nUnlike in the 1D case, the asymptotics contains a time independent term. The\nderivation of its spatial asymptotics is the technical core of the paper.\nBesides numerical applications, the obtained results play a crucial role in the\nanalysis of spatio-temporal patterns for reaction-diffusion equations on\nlattices, in particular rattling patterns for hysteretic diffusion systems.\n", " Recently there has been renewed interest in signed distance bound\nrepresentations due to their unique properties for 3D shape modelling. This is\nespecially the case for deep learning-based bounds. However, it is beneficial\nto work with polygons in most computer-graphics applications. Thus, in this\npaper we introduce and investigate an asymptotically fast method for\ntransforming signed distance bounds into polygon meshes. This is achieved by\ncombining the principles of sphere tracing (or ray marching) with traditional\npolygonization techniques, such as Marching Cubes. We provide theoretical and\nexperimental evidence that this approach is of the $O(N^2\\log N)$ computational\ncomplexity for a polygonization grid with $N^3$ cells. The algorithm is tested\non both a set of primitive shapes as well as signed distance bounds generated\nfrom point clouds by machine learning (and represented as neural networks).\nGiven its speed, implementation simplicity and portability, we argue that it\ncould prove useful during the modelling stage as well as in shape compression\nfor storage.\n The code is available here: https://github.com/nenadmarkus/gridhopping\n", " Implicit Neural Representations (INR) or neural fields have emerged as a\npopular framework to encode multimedia signals such as images and radiance\nfields while retaining high-quality. Recently, learnable feature grids proposed\nby Instant-NGP have allowed significant speed-up in the training as well as the\nsampling of INRs by replacing a large neural network with a multi-resolution\nlook-up table of feature vectors and a much smaller neural network. However,\nthese feature grids come at the expense of large memory consumption which can\nbe a bottleneck for storage and streaming applications. In this work, we\npropose SHACIRA, a simple yet effective task-agnostic framework for compressing\nsuch feature grids with no additional post-hoc pruning/quantization stages. We\nreparameterize feature grids with quantized latent weights and apply entropy\nregularization in the latent space to achieve high levels of compression across\nvarious domains. Quantitative and qualitative results on diverse datasets\nconsisting of images, videos, and radiance fields, show that our approach\noutperforms existing INR approaches without the need for any large datasets or\ndomain-specific heuristics. Our project page is available at\nhttp://shacira.github.io .\n", " The popularity and widespread use of pruning and quantization is driven by\nthe severe resource constraints of deploying deep neural networks to\nenvironments with strict latency, memory and energy requirements. These\ntechniques achieve high levels of compression with negligible impact on\ntop-line metrics (top-1 and top-5 accuracy). However, overall accuracy hides\ndisproportionately high errors on a small subset of examples; we call this\nsubset Compression Identified Exemplars (CIE). We further establish that for\nCIE examples, compression amplifies existing algorithmic bias. Pruning\ndisproportionately impacts performance on underrepresented features, which\noften coincides with considerations of fairness. Given that CIE is a relatively\nsmall subset but a great contributor of error in the model, we propose its use\nas a human-in-the-loop auditing tool to surface a tractable subset of the\ndataset for further inspection or annotation by a domain expert. We provide\nqualitative and quantitative support that CIE surfaces the most challenging\nexamples in the data distribution for human-in-the-loop auditing.\n", " We study asymptotic distribution of eigen-values $\\omega$ of a quadratic\noperator polynomial of the following form $(\\omega^2-L(\\omega))\\phi_\\omega=0$,\nwhere $L(\\omega)$ is a second order differential positive elliptic operator\nwith quadratic dependence on the spectral parameter $\\omega$. We derive\nasymptotics of the spectral density in this problem and show how to compute\ncoefficients of its asymptotic expansion from coefficients of the asymptotic\nexpansion of the trace of the heat kernel of $L(\\omega)$. The leading term in\nthe spectral asymptotics is the same as for a Laplacian in a cavity. The\nresults have a number of physical applications. We illustrate them by examples\nof field equations in external stationary gravitational and gauge backgrounds.\n", " Popular peer to peer streaming media systems such as PPLive and UUSee rely on\nperiodic buffer-map exchange between peers for proper operation. The buffer-map\nexchange contains redundant information which causes non-negligible overhead.\nIn this paper we present a theoretical framework to study how the overhead can\nbe lowered. Differentiating from the traditional data compression approach, we\ndo not treat each buffer-map as an isolated data block, but consider the\ncorrelations between the sequentially exchanged buffer-maps. Under this\nframework, two buffer-map compression schemes are proposed and the correctness\nof the schemes is proved mathematically. Moreover, we derive the theoretical\nlimit of compression gain based on probability theory and information theory.\nBased on the system parameters of UUSee (a popular P2P streaming platform), our\nsimulations show that the buffer-map sizes are reduced by 86% and 90% (from 456\nbits down to only 66 bits and 46 bits) respectively after applying our schemes.\nFurthermore, by combining with the traditional compression methods (on\nindividual blocks), the sizes are decreased by 91% and 95% (to 42 bits and 24\nbits) respectively. Our study provides a guideline for developing practical\ncompression algorithms.\n", " We consider two specific thermodynamic cycles of engine operating in a finite\ntime coupled to two thermal reservoirs with a finite heat capacity: The\nCarnot-type cycle and the Lorenz-type cycle. By means of the endo-reversible\nthermodynamics, we then discuss the power output of engine and its\noptimization. In doing so, we treat the temporal duration of a single cycle\nrigorously, i.e., without neglecting the duration of its adiabatic parts. Then\nwe find that the maximally obtainable power output P_m and the engine\nefficiency \\eta_m at the point of P_m explicitly depend on the heat conductance\nand the compression ratio. From this, it is immediate to observe that the\nwell-known results available in many references, in particular the\n(compression-ratio-independent) Curzon-Ahlborn-Novikov expressions such as\n\\eta_m --> \\eta_{CAN} = 1 - (T_L/T_H)^(1/2) with the temperatures (T_H, T_L) of\nhot and cold reservoirs only, can be recovered, but, significantly enough, in\nthe limit of a vanishingly small heat conductance and an infinitely large\ncompression ratio only. Our result implies that the endo-reversible model of a\nthermal machine operating in a finite time and so producing a finite power\noutput with the Curzon-Ahlborn-Novikov results should be limited in its own\nvalidity regime.\n", " Implicit neural representations (INRs) recently achieved great success in\nimage representation and compression, offering high visual quality and fast\nrendering speeds with 10-1000 FPS, assuming sufficient GPU resources are\navailable. However, this requirement often hinders their use on low-end devices\nwith limited memory. In response, we propose a groundbreaking paradigm of image\nrepresentation and compression by 2D Gaussian Splatting, named GaussianImage.\nWe first introduce 2D Gaussian to represent the image, where each Gaussian has\n8 parameters including position, covariance and color. Subsequently, we unveil\na novel rendering algorithm based on accumulated summation. Remarkably, our\nmethod with a minimum of 3$\\times$ lower GPU memory usage and 5$\\times$ faster\nfitting time not only rivals INRs (e.g., WIRE, I-NGP) in representation\nperformance, but also delivers a faster rendering speed of 1500-2000 FPS\nregardless of parameter size. Furthermore, we integrate existing vector\nquantization technique to build an image codec. Experimental results\ndemonstrate that our codec attains rate-distortion performance comparable to\ncompression-based INRs such as COIN and COIN++, while facilitating decoding\nspeeds of approximately 1000 FPS. Additionally, preliminary proof of concept\nshows that our codec surpasses COIN and COIN++ in performance when using\npartial bits-back coding. Code will be available at\nhttps://github.com/Xinjie-Q/GaussianImage.\n", " We study the critical probability for the metastable phase transition of the\ntwo-dimensional anisotropic bootstrap percolation model with\n$(1,2)$-neighbourhood and threshold $r = 3$. The first order asymptotics for\nthe critical probability were recently determined by the first and second\nauthors. Here we determine the following sharp second and third order\nasymptotics: \\[\n p_c\\big( [L]^2,\\mathcal{N}_{(1,2)},3 \\big) \\; = \\; \\frac{(\\log \\log\nL)^2}{12\\log L} \\, - \\, \\frac{\\log \\log L \\, \\log \\log \\log L}{ 3\\log L}\n + \\frac{\\left(\\log \\frac{9}{2} + 1 \\pm o(1) \\right)\\log \\log L}{6\\log L}. \\]\nWe note that the second and third order terms are so large that the first order\nasymptotics fail to approximate $p_c$ even for lattices of size well beyond\n$10^{10^{1000}}$.\n", " The authors analyze the asymptotics of eigenvalues of Toeplitz matrices with\ncertain continuous and discontinuous symbols. In particular, the authors prove\na conjecture of Levitin and Shargorodsky on the near-periodicity of Toeplitz\neigenvalues.\n", " The use of large transformer-based models such as BERT, GPT, and T5 has led\nto significant advancements in natural language processing. However, these\nmodels are computationally expensive, necessitating model compression\ntechniques that reduce their size and complexity while maintaining accuracy.\nThis project investigates and applies knowledge distillation for BERT model\ncompression, specifically focusing on the TinyBERT student model. We explore\nvarious techniques to improve knowledge distillation, including experimentation\nwith loss functions, transformer layer mapping methods, and tuning the weights\nof attention and representation loss and evaluate our proposed techniques on a\nselection of downstream tasks from the GLUE benchmark. The goal of this work is\nto improve the efficiency and effectiveness of knowledge distillation, enabling\nthe development of more efficient and accurate models for a range of natural\nlanguage processing tasks.\n", " We study the convolution function $$ C[f(x)] := \\int_1^x f(y)f({x\\over y})\n{{\\rm d} y\\over y} $$ when $f(x)$ is a suitable number-theoretic error term.\nAsymptotics and upper bounds for $C[f(x)]$ are derived from mean square bounds\nfor $f(x)$. Some applications are given, in particular to\n$|\\zeta(1/2+ix)|^{2k}$ and the classical Rankin--Selberg problem from analytic\nnumber theory.\n", " We study asymptotics of generalized eigenvectors associated with Jacobi\nmatrices. Under weak conditions on the coefficients we identify when the\nmatrices are self-adjoint and show that they satisfy strong non-subordinacy\ncondition.\n", " Diffusion models have emerged as a powerful new family of deep generative\nmodels with record-breaking performance in many applications, including image\nsynthesis, video generation, and molecule design. In this survey, we provide an\noverview of the rapidly expanding body of work on diffusion models,\ncategorizing the research into three key areas: efficient sampling, improved\nlikelihood estimation, and handling data with special structures. We also\ndiscuss the potential for combining diffusion models with other generative\nmodels for enhanced results. We further review the wide-ranging applications of\ndiffusion models in fields spanning from computer vision, natural language\ngeneration, temporal data modeling, to interdisciplinary applications in other\nscientific disciplines. This survey aims to provide a contextualized, in-depth\nlook at the state of diffusion models, identifying the key areas of focus and\npointing to potential areas for further exploration. Github:\nhttps://github.com/YangLing0818/Diffusion-Models-Papers-Survey-Taxonomy.\n", " In this work we study the Plancherel-Rotach type asymptotics for\nStieltjes-Wigert orthogonal polynomials with complex scaling. The main term of\nthe asymptotics contains Ramanujan function $A_{q}(z)$ for the scaling\nparameter on the vertical line $\\Re(s)=2$, while the main term of the\nasymptotics involves the theta functions for the scaling parameter in the strip\n$0<\\Re(s)<2$. In the latter case the number theoretical property of the scaling\nparameter completely determines the order of the error term. $ $ These\nasymptotic formulas may provide some insights to some new random matrix model\nand also add a new link between special functions and number theory.\n", " We give precise asymptotics to the number of random walks in the standard\northogonal lattice in $\\mathbb{R}^d$ that return to the starting point at step\n$2n$, both for all such walks and for the ones that return for the first time.\nThe first set of asymptotics is obtained in an elementary way, by using a\ncombinatorial and geometric multiplication principle together with the\nclassical theory of Legendre polynomials. As an easy consequence we obtain a\nunified proof of P\\'{o}lya's theorem. By showing that the relevant generating\nfunctions are $\\Delta$-analytic, we use the deeper Tauberian theory of\nsingularity analysis to obtain the asymptotics for the first return paths.\n", " We study the problem of representing all distances between $n$ points in\n$\\mathbb R^d$, with arbitrarily small distortion, using as few bits as\npossible. We give asymptotically tight bounds for this problem, for Euclidean\nmetrics, for $\\ell_1$ (a.k.a.~Manhattan) metrics, and for general metrics.\n Our bounds for Euclidean metrics mark the first improvement over compression\nschemes based on discretizing the classical dimensionality reduction theorem of\nJohnson and Lindenstrauss (Contemp.~Math.~1984). Since it is known that no\nbetter dimension reduction is possible, our results establish that Euclidean\nmetric compression is possible beyond dimension reduction.\n", " Thermal conductivity measurements over variable lengths on nanostructures\nsuch as nanowires provide important information about the mean free paths\n(MFPs) of the phonons responsible for heat conduction. However, nearly all of\nthese measurements have been interpreted using an average MFP even though\nphonons in many crystals possess a broad MFP spectrum. Here, we present a\nreconstruction method to obtain MFP spectra of nanostructures from\nvariable-length thermal conductivity measurements. Using this method, we\ninvestigate recently reported length-dependent thermal conductivity\nmeasurements on SiGe alloy nanowires and suspended graphene ribbons. We find\nthat the recent measurements on graphene imply that 70 % of the heat in\ngraphene is carried by phonons with MFPs longer than 1 micron.\n", " The 0-instanton solution of Painlev\\'e I is a sequence $(u_{n,0})$ of complex\nnumbers which appears universally in many enumerative problems in algebraic\ngeometry, graph theory, matrix models and 2-dimensional quantum gravity. The\nasymptotics of the 0-instanton $(u_{n,0})$ for large $n$ were obtained by the\nthird author using the Riemann-Hilbert approach. For $k=0,1,2,...$, the\n$k$-instanton solution of Painlev\\'e I is a doubly-indexed sequence $(u_{n,k})$\nof complex numbers that satisfies an explicit quadratic non-linear recursion\nrelation. The goal of the paper is three-fold: (a) to compute the asymptotics\nof the 1-instanton sequence $(u_{n,1})$ to all orders in $1/n$ by using the\nRiemann-Hilbert method, (b) to present formulas for the asymptotics of\n$(u_{n,k})$ for fixed $k$ and to all orders in $1/n$ using resurgent analysis,\nand (c) to confirm numerically the predictions of resurgent analysis. We point\nout that the instanton solutions display a new type of Stokes behavior, induced\nfrom the tritronqu\\'ee Painlev\\'e transcendents, and which we call the induced\nStokes phenomenon. The asymptotics of the 2-instanton and beyond exhibits new\nphenomena not seen in 0 and 1-instantons, and their enumerative context is at\npresent unknown.\n", " In this paper, we consider the problem of improving the inference latency of\nlanguage model-based dense retrieval systems by introducing structural\ncompression and model size asymmetry between the context and query encoders.\nFirst, we investigate the impact of pre and post-training compression on the\nMSMARCO, Natural Questions, TriviaQA, SQUAD, and SCIFACT, finding that\nasymmetry in the dual encoders in dense retrieval can lead to improved\ninference efficiency. Knowing this, we introduce Kullback Leibler Alignment of\nEmbeddings (KALE), an efficient and accurate method for increasing the\ninference efficiency of dense retrieval methods by pruning and aligning the\nquery encoder after training. Specifically, KALE extends traditional Knowledge\nDistillation after bi-encoder training, allowing for effective query encoder\ncompression without full retraining or index generation. Using KALE and\nasymmetric training, we can generate models which exceed the performance of\nDistilBERT despite having 3x faster inference.\n", " We develop an efficient algorithm, coordinate descent FCI (CDFCI), for the\nelectronic structure ground state calculation in the configuration interaction\nframework. CDFCI solves an unconstrained non-convex optimization problem, which\nis a reformulation of the full configuration interaction eigenvalue problem,\nvia an adaptive coordinate descent method with a deterministic compression\nstrategy. CDFCI captures and updates appreciative determinants with different\nfrequencies proportional to their importance. We show that CDFCI produces\naccurate variational energy for both static and dynamic correlation by\nbenchmarking the binding curve of nitrogen dimer in the cc-pVDZ basis with\n$10^{-3}$ mHa accuracy. We also demonstrate the efficiency and accuracy of\nCDFCI for strongly correlated chromium dimer in the Ahlrichs VDZ basis and\nproduces state-of-the-art variational energy.\n", " For a car manufacturer, reducing the weight of vehicles is an obvious aim.\nReplacing steel by aluminium moves towards that goal. Unfortunately,\naluminium's stamping numerical simulation results are not yet as reliable as\nthose of steel. Punch-strength and spring-back phenomena are not correctly\ndescribed. This study on aluminium validates the behaviour law Hill 48\nquadratic yield criterion with both isotropic and kinematic hardening. It is\nbased on the yield surface and on associated experimental tests (uniaxial test,\nplane tensile test, plane compression and tensile shearing).\n", " We construct a canonical differential structure on the configuration space\n$\\Upsilon$ over a singular base space $X$ and with a general invariant measure\n$\\mu$ on $\\Upsilon$. We present an analytic structure on $\\Upsilon$,\nconstructing a strongly local Dirichlet form $\\mathcal E$ on $L^2(\\Upsilon,\n\\mu)$ for $\\mu$ in a large class of probability measures. We then investigate\nthe geometric structure of the extended metric measure space $\\Upsilon$ endowed\nwith the $L^2$-transportation extended distance $\\mathsf{d}_{\\Upsilon}$ and\nwith the measure $\\mu$. By establishing Rademacher- and\nSobolev-to-Lipschitz-type properties for $\\mathcal E$, we finally provide a\ncomplete identification of the analytic and the geometric structure -- the\ncanonical differential structure induced on $\\Upsilon$ by $X$ and $\\mu$ --\nshowing that $\\mathcal E$ coincides with the Cheeger energy of\n$(\\Upsilon,\\mathsf{d}_{\\Upsilon},\\mu)$ and that the intrinsic distance of\n$\\mathcal E$ coincides with $\\mathsf{d}_{\\Upsilon}$. The class of base spaces\nto which our results apply includes sub-Riemannian manifolds, RCD spaces, and\npath/loop spaces over Riemannian manifolds; as for $\\mu$ our results include\nquasi-Gibbs measures, in particular: Poisson measures, canonical Gibbs\nmeasures, as well as some determinantal/permanental point processes\n(sine$_\\beta$, Airy$_\\beta$, Bessel$_{\\alpha,\\beta}$, Ginibre). A number of\napplications to interacting particle systems and infinite-dimensional metric\nmeasure geometry are also discussed. In particular, we prove the universality\nof the $L^2$-transportation distance $\\mathsf{d}_{\\Upsilon}$ for the Varadhan\nshort-time asymptotics for diffusions on $\\Upsilon$, regardless of the choice\nof $\\mu$. Many of our results are new even in the case of configuration spaces\nover Euclidean spaces.\n", " In this paper, we investigate optimal linear approximations\n($n$-approximation numbers ) of the embeddings from the Sobolev spaces $H^r\\\n(r>0)$ for various equivalent norms and the Gevrey type spaces\n$G^{\\alpha,\\beta}\\ (\\alpha,\\beta>0)$ on the sphere $\\Bbb S^d$ and on the ball\n$\\Bbb B^d$, where the approximation error is measured in the $L_2$-norm. We\nobtain preasymptotics, asymptotics, and strong equivalences of the above\napproximation numbers as a function in $n$ and the dimension $d$. We emphasis\nthat all equivalence constants in the above preasymptotics and asymptotics are\nindependent of the dimension $d$ and $n$. As a consequence we obtain that for\nthe absolute error criterion the approximation problems $I_d: H^{r}\\to L_2$ are\nweakly tractable if and only if $r>1$, not uniformly weakly tractable, and do\nnot suffer from the curse of dimensionality. We also prove that for any\n$\\alpha,\\beta>0$, the approximation problems $I_d: G^{\\alpha,\\beta}\\to L_2$ are\nuniformly weakly tractable, not polynomially tractable, and quasi-polynomially\ntractable if and only if $\\alpha\\ge 1$.\n", " In this study, we investigate whether attention-based information flow inside\nlarge language models (LLMs) is aggregated through noticeable patterns for long\ncontext processing. Our observations reveal that LLMs aggregate information\nthrough Pyramidal Information Funneling where attention is scattering widely in\nlower layers, progressively consolidating within specific contexts, and\nultimately focusin on critical tokens (a.k.a massive activation or attention\nsink) in higher layers. Motivated by these insights, we developed PyramidKV, a\nnovel and effective KV cache compression method. This approach dynamically\nadjusts the KV cache size across different layers, allocating more cache in\nlower layers and less in higher ones, diverging from traditional methods that\nmaintain a uniform KV cache size. Our experimental evaluations, utilizing the\nLongBench benchmark, show that PyramidKV matches the performance of models with\na full KV cache while retaining only 12% of the KV cache, thus significantly\nreducing memory usage. In scenarios emphasizing memory efficiency, where only\n0.7% of the KV cache is maintained, PyramidKV surpasses other KV cache\ncompression techniques achieving up to a 20.5 absolute accuracy improvement on\nTREC.\n", " Prediction+optimization is a common real-world paradigm where we have to\npredict problem parameters before solving the optimization problem. However,\nthe criteria by which the prediction model is trained are often inconsistent\nwith the goal of the downstream optimization problem. Recently,\ndecision-focused prediction approaches, such as SPO+ and direct optimization,\nhave been proposed to fill this gap. However, they cannot directly handle the\nsoft constraints with the $max$ operator required in many real-world\nobjectives. This paper proposes a novel analytically differentiable surrogate\nobjective framework for real-world linear and semi-definite negative quadratic\nprogramming problems with soft linear and non-negative hard constraints. This\nframework gives the theoretical bounds on constraints' multipliers, and derives\nthe closed-form solution with respect to predictive parameters and thus\ngradients for any variable in the problem. We evaluate our method in three\napplications extended with soft constraints: synthetic linear programming,\nportfolio optimization, and resource provisioning, demonstrating that our\nmethod outperforms traditional two-staged methods and other decision-focused\napproaches.\n", " This article is centered around generalizing a previous implicit function\ntheorem of the author to be applicable for maps f:E sqcap F to F which can be\nlifted to Keller C^k_pi maps f_i:E sqcap F_i to F_i with F_i Banach and\nF=projlim F_i . We prove theorems about existence and differentiability of\nfunctions g satisfying f(x,g(x))=b constant. In addition to these abstract\ntheorems, we give several examples of applications to proving smooth dependence\nof the solution on initial/boundary values and the nonlinearity in nonlinear\n(partial) differential equations.\n", " The discovery of close-to-star gas-giant exo-planets lends support to the\nidea of Earth's origin as a Jupiter-like gas-giant and to the consequences of\nits compression, including whole-Earth decompression dynamics that gives rise,\nwithout requiring mantle convection, to the myriad measurements and\nobservations whose descriptions are attributed to plate tectonics. I propose\nhere another, unanticipated consequence of whole-Earth decompression dynamics:\nnamely, a specific, dominant, non-erosion, underlying initiation-mechanism\nprecursor for submarine canyons that follows as a direct consequence of Earth's\nearly origin as a Jupiter-like gas-giant.\n", " We present a generalization of the Fefferman-de la Llave decomposition of the\nCoulomb potential to quite arbitrary radial functions $V$ on $R^n$ going to\nzero at infinity. This generalized decomposition can be used to extend previous\nresults on $N$-body quantum systems with Coulomb interaction to a more general\nclass of interactions. As an example of such an application we derive the high\ndensity asymptotics of the ground state energy of jellium with Yukawa\ninteraction in the thermodynamic limit, using a correlation estimate by Graf\nand Solovej.\n", " The Divider set, as an innovative alternative concept to maximal disks,\nVoronoi sets and cut loci, is presented with a formal definition based on\ntopology and differential geometry. The relevant mathematical theory by\nprevious authors and a comparison with other medial axis definitions is\npresented. Appropriate applications are proposed and examined.\n", " Resurgent-transseries solutions to Painleve equations may be recursively\nconstructed out of these nonlinear differential-equations -- but require Stokes\ndata to be globally defined over the complex plane. Stokes data explicitly\nconstruct connection-formulae which describe the nonlinear Stokes phenomena\nassociated to these solutions, via implementation of Stokes transitions acting\non the transseries. Nonlinear resurgent Stokes data lack, however, a\nfirst-principle computational approach, hence are hard to determine\ngenerically. In the Painleve I and Painleve II contexts, nonlinear Stokes data\nget further hindered as these equations are resonant, with non-trivial\nconsequences for the interconnections between transseries sectors, bridge\nequations, and associated Stokes coefficients. In parallel to this, the\nPainleve I and Painleve II equations are string-equations for two-dimensional\nquantum (super) gravity and minimal string theories, where Stokes data have\nnatural ZZ-brane interpretations. This work computes for the first time the\ncomplete, analytical, resurgent Stokes data for the first two Painleve\nequations, alongside their quantum gravity or minimal string incarnations. The\nmethod developed herein, dubbed \"closed-form asymptotics\", makes sole use of\nresurgent large-order asymptotics of transseries solutions -- alongside a\ncareful analysis of the role resonance plays. Given its generality, it may be\napplicable to other distinct (nonlinear, resonant) problems. Results for\nanalytical Stokes coefficients have natural structures, which are described,\nand extensive high-precision numerical tests corroborate all analytical\npredictions. Connection-formulae are explicitly constructed, with rather simple\nand compact final results encoding the full Stokes data, and further allowing\nfor exact monodromy checks -- hence for an analytical proof of our results.\n", " In the realm of deep learning, the self-attention mechanism has substantiated\nits pivotal role across a myriad of tasks, encompassing natural language\nprocessing and computer vision. Despite achieving success across diverse\napplications, the traditional self-attention mechanism primarily leverages\nlinear transformations for the computation of query, key, and value (QKV),\nwhich may not invariably be the optimal choice under specific circumstances.\nThis paper probes into a novel methodology for QKV computation-implementing a\nspecially-designed neural network structure for the calculation. Utilizing a\nmodified Marian model, we conducted experiments on the IWSLT 2017\nGerman-English translation task dataset and juxtaposed our method with the\nconventional approach. The experimental results unveil a significant\nenhancement in BLEU scores with our method. Furthermore, our approach also\nmanifested superiority when training the Roberta model with the Wikitext-103\ndataset, reflecting a notable reduction in model perplexity compared to its\noriginal counterpart. These experimental outcomes not only validate the\nefficacy of our method but also reveal the immense potential in optimizing the\nself-attention mechanism through neural network-based QKV computation, paving\nthe way for future research and practical applications. The source code and\nimplementation details for our proposed method can be accessed at\nhttps://github.com/ocislyjrti/NeuralAttention.\n", " Comparative analysis of scalar fields is an important problem with various\napplications including feature-directed visualization and feature tracking in\ntime-varying data. Comparing topological structures that are abstract and\nsuccinct representations of the scalar fields lead to faster and meaningful\ncomparison. While there are many distance or similarity measures to compare\ntopological structures in a global context, there are no known measures for\ncomparing topological structures locally. While the global measures have many\napplications, they do not directly lend themselves to fine-grained analysis\nacross multiple scales. We define a local variant of the tree edit distance and\napply it towards local comparative analysis of merge trees with support for\nfiner analysis. We also present experimental results on time-varying scalar\nfields, 3D cryo-electron microscopy data, and other synthetic data sets to show\nthe utility of this approach in applications like symmetry detection and\nfeature tracking.\n", " We derive uniform and non-uniform asymptotics of the Charlier polynomials by\nusing difference equation methods alone. The Charlier polynomials are special\nin that they do not fit into the framework of the turning point theory, despite\nthe fact that they are crucial in the Askey scheme. In this paper, asymptotic\napproximations are obtained respectively in the outside region, an intermediate\nregion, and near the turning points. In particular, we obtain uniform\nasymptotic approximation at a pair of coalescing turning points with the aid of\na local transformation. We also give a uniform approximation at the origin by\napplying the method of dominant balance and several matching techniques.\n", " This work investigates spectrum and root functions (that is, eigen- and\nassociated functions) of a Sturm-Liouville problem involving an abstract linear\noperator (nonselfadjoint in general) in the equation together with\nsupplementary transmission conditions at the one interior singular point. So,\nthe problem under consideration is not pure differential pro lem. At first we\nestablish isomorphism and coerciveness with respect to the spectral parameter\nfor the corresponding nonhomogeneous problem. Then by suggesting an own our\nmethod we prove that the spectrum of the considered problem is discrete and\nderive an asymptotic approximation formulas for the eigenvalues. We must note\nthat Asymptotics of the eigenvalues of such type problems is investigated at\nfirst in literature in the present work and the obtained results are new even\nin the continuous case (i.e. without transmission conditions). Finally it is\nshown that the system of root functions form an Abel basis of the corresponding\nHilbert space.\n", " We consider the problem of counting real ramified covers of\n$\\mathbb{C}\\mathbb{P}^1$ by genus $g$ real Riemann surfaces with ramification\nprofiles $\\lambda$, $\\mu$ over $0$, $\\infty$ respectively, and with\nramification $(3,1,\\ldots,1)$ over other real branch points. The resulted\nanswers are real double Hurwitz numbers with $3$-cycles. We compute real double\nHurwitz numbers with $3$-cycles via tropical covers, and introduce a lower\nbound, which does not depend on the distributions of real branch points, of\nthese numbers. We give a non-vanishing theorem for the invariant. When the\ninvariant is non-zero, we obtain a lower bound of a logarithmic asymptotics for\nthe invariants as the degree tends to infinity. It implies the logarithmic\nasymptotic growth of real double Hurwitz numbers with $3$-cycles.\n", " It is highly desirable to combine recent advances in the topological quantum\nphases with technologically relevant materials. Chromium dioxide (CrO2) is a\nhalf-metallic material, widely used in high-end data storage applications.\nUsing first principles calculations, we show via interfacial orbital design\nthat a novel class of half semi-metallic Dirac electronic phase emerges at the\ninterface CrO2 with TiO2 in both thin film and superlattice configurations,\nwith four spin-polarized Dirac points in momentum-space (k-space) band\nstructure. When the spin and orbital degrees of freedom are allowed to couple,\nthe CrO2/TiO2 superlattice becomes a Chern insulator without external fields or\nadditional doping. With topological gaps equivalent to 43 Kelvin and a Chern\nnumber 2, the ensuing quantization of Hall conductance to 2e^2/h will enable\npotential development of these highly industrialized oxides for applications in\ntopologically high fidelity data storage and energy-efficient electronic and\nspintronic devices.\n", " This paper tackles a new problem setting: reinforcement learning with\npixel-wise rewards (pixelRL) for image processing. After the introduction of\nthe deep Q-network, deep RL has been achieving great success. However, the\napplications of deep RL for image processing are still limited. Therefore, we\nextend deep RL to pixelRL for various image processing applications. In\npixelRL, each pixel has an agent, and the agent changes the pixel value by\ntaking an action. We also propose an effective learning method for pixelRL that\nsignificantly improves the performance by considering not only the future\nstates of the own pixel but also those of the neighbor pixels. The proposed\nmethod can be applied to some image processing tasks that require pixel-wise\nmanipulations, where deep RL has never been applied. We apply the proposed\nmethod to three image processing tasks: image denoising, image restoration, and\nlocal color enhancement. Our experimental results demonstrate that the proposed\nmethod achieves comparable or better performance, compared with the\nstate-of-the-art methods based on supervised learning.\n", " This paper studies the first moment of symmetric-square $L$-functions at the\ncritical point in the weight aspect. Asymptotics with the best known error term\n$O(k^{-1/2})$ were obtained independently by Fomenko in 2005 and by Sun in\n2013. We prove that there is an extra main term of size $k^{-1/2}$ in the\nasymptotic formula and show that the remainder term decays exponentially in\n$k$. The twisted first moment was evaluated asymptotically by Ng Ming Ho with\nthe error bounded by $lk^{-1/2+\\epsilon}$. We improve the error bound to\n$l^{5/6+\\epsilon}k^{-1/2+\\epsilon}$ unconditionally and to\n$l^{1/2+\\epsilon}k^{-1/2}$ under the Lindel\\\"{o}f hypothesis for quadratic\nDirichlet $L$-functions.\n", " Let $n$ and $k$ be nonnegative integers such that $1\\le k\\le n+1$. The convex\ncone $\\mathcal{F}_+^{k:n}$ of all functions $f$ on an arbitrary interval\n$I\\subseteq\\mathbb{R}$ whose derivatives $f^{(j)}$ of orders $j=k-1,\\dots,n$\nare nondecreasing is characterized in terms of extreme rays of the cone\n$\\mathcal{F}_+^{k:n}$. A simple description of the convex cone dual to\n$\\mathcal{F}_+^{k:n}$ is given. These results are useful in, and were motivated\nby, applications in probability. In fact, the results are obtained in a more\ngeneral setting with certain generalized derivatives of $f$ of the $j$th order\nin place of $f^{(j)}$. Somewhat similar results were previously obtained in the\ncase when the left endpoint of the interval $I$ is finite, with certain\nadditional integrability conditions; such conditions fail to hold in the\nmentioned applications.\n", " I postulate that human or other intelligent agents function or should\nfunction as follows. They store all sensory observations as they come - the\ndata is holy. At any time, given some agent's current coding capabilities, part\nof the data is compressible by a short and hopefully fast program / description\n/ explanation / world model. In the agent's subjective eyes, such data is more\nregular and more \"beautiful\" than other data. It is well-known that knowledge\nof regularity and repeatability may improve the agent's ability to plan actions\nleading to external rewards. In absence of such rewards, however, known beauty\nis boring. Then \"interestingness\" becomes the first derivative of subjective\nbeauty: as the learning agent improves its compression algorithm, formerly\napparently random data parts become subjectively more regular and beautiful.\nSuch progress in compressibility is measured and maximized by the curiosity\ndrive: create action sequences that extend the observation history and yield\npreviously unknown / unpredictable but quickly learnable algorithmic\nregularity. We discuss how all of the above can be naturally implemented on\ncomputers, through an extension of passive unsupervised learning to the case of\nactive data selection: we reward a general reinforcement learner (with access\nto the adaptive compressor) for actions that improve the subjective\ncompressibility of the growing data. An unusually large breakthrough in\ncompressibility deserves the name \"discovery\". The \"creativity\" of artists,\ndancers, musicians, pure mathematicians can be viewed as a by-product of this\nprinciple. Several qualitative examples support this hypothesis.\n", " We consider the thin-film equation for a class of free boundary conditions\nmodelling friction at the contact line, as introduced by E and Ren. Our\nanalysis focuses on formal long-time asymptotics of solutions in the perfect\nwetting regime. In particular, through the analysis of quasi-self-similar\nsolutions, we characterize the profile and the spreading rate of solutions\ndepending on the strength of friction at the contact line, as well as their\n(global or local) corrections, which are due to the dynamical nature of the\nfree boundary conditions. These results are complemented with full transient\nnumerical solutions of the free boundary problem.\n", " In this article we present a class of codes with few weights arising from\nspecial type of linear sets. We explicitly show the weights of such codes,\ntheir weight enumerator and possible choices for their generator matrices. In\nparticular, our construction yields also to linear codes with three weights\nand, in some cases, to almost MDS codes. The interest for these codes relies on\ntheir applications to authentication codes and secret schemes, and their\nconnections with further objects such as association schemes and graphs.\n", " We present a particular class of solutions in Cartesian, cylindrical and\nspherical coordinates of the non-dispersive travelling wave variety that\npropagate an envelope of varying vorticity some of which include topological\nwaves with parallel electric and magnetic components. The significance of these\nsolutions is examined in the recently proposed Axion-Maxwell field theory with\npotential applications in material science and topological insulators.\n", " We consider Markovian models on graphs with local dynamics. We show that,\nunder suitable conditions, such Markov chains exhibit both rapid convergence to\nequilibrium and strong concentration of measure in the stationary distribution.\nWe illustrate our results with applications to some known chains from computer\nscience and statistical mechanics.\n", " We develop a theory of discrete directional Gabor frames for functions\ndefined on the $d$-dimensional Euclidean space. Our construction incorporates\nthe concept of ridge functions into the theory of isotropic Gabor systems, in\norder to develop an anisotropic Gabor system with strong directional\nsensitivity. We present sufficient conditions on a window function $g$ and a\nsampling set $\\Lambda$ for the corresponding directional Gabor system\n$\\{g_{m,t,u}\\}_{(m,t,u)\\in\\Lambda}$ to form a discrete frame. Explicit\nestimates on the frame bounds are developed. A numerical implementation of our\nscheme is also presented, and is shown to perform competitively in compression\nand denoising against state-of-the-art anisotropic methods.\n", " Convolutional Neural Networks (CNN) are the state-of-the-art in the field of\nvisual computing. However, a major problem with CNNs is the large number of\nfloating point operations (FLOPs) required to perform convolutions for large\ninputs. When considering the application of CNNs to video data, convolutional\nfilters become even more complex due to the extra temporal dimension. This\nleads to problems when respective applications are to be deployed on mobile\ndevices, such as smart phones, tablets, micro-controllers or similar,\nindicating less computational power.\n Kim et al. (2016) proposed using a Tucker-decomposition to compress the\nconvolutional kernel of a pre-trained network for images in order to reduce the\ncomplexity of the network, i.e. the number of FLOPs. In this paper, we\ngeneralize the aforementioned method for application to videos (and other 3D\nsignals) and evaluate the proposed method on a modified version of the THETIS\ndata set, which contains videos of individuals performing tennis shots. We show\nthat the compressed network reaches comparable accuracy, while indicating a\nmemory compression by a factor of 51. However, the actual computational\nspeed-up (factor 1.4) does not meet our theoretically derived expectation\n(factor 6).\n" ]

[ " We consider two Lax systems for the homogeneous Painlev\\'{e} II equation: one\nof size $2\\times 2$ studied by Flaschka and Newell in the early 1980's, and one\nof size $4\\times 4$ introduced by Delvaux-Kuijlaars-Zhang and Duits-Geudens in\nthe early 2010's. We prove that solutions to the $4\\times 4$ system can be\nderived from those to the $2\\times 2$ system via an integral transform, and\nconsequently relate the Stokes multipliers for the two systems. As corollaries\nwe are able to express two kernels for determinantal processes as contour\nintegrals involving the Flaschka-Newell Lax system: the tacnode kernel arising\nin models of nonintersecting paths, and a critical kernel arising in a\ntwo-matrix model.\n", " We study the completion of approximately low rank matrices with entries\nmissing not at random (MNAR). In the context of typical large-dimensional\nstatistical settings, we establish a framework for the performance analysis of\nthe nuclear norm minimization ($\\ell_1^*$) algorithm. Our framework produces\n\\emph{exact} estimates of the worst-case residual root mean squared error and\nthe associated phase transitions (PT), with both exhibiting remarkably simple\ncharacterizations. Our results enable to {\\it precisely} quantify the impact of\nkey system parameters, including data heterogeneity, size of the missing block,\nand deviation from ideal low rankness, on the accuracy of $\\ell_1^*$-based\nmatrix completion. To validate our theoretical worst-case RMSE estimates, we\nconduct numerical simulations, demonstrating close agreement with their\nnumerical counterparts.\n", " Confidence calibration is central to providing accurate and interpretable\nuncertainty estimates, especially under safety-critical scenarios. However, we\nfind that existing calibration algorithms often overlook the issue of\n*proximity bias*, a phenomenon where models tend to be more overconfident in\nlow proximity data (i.e., data lying in the sparse region of the data\ndistribution) compared to high proximity samples, and thus suffer from\ninconsistent miscalibration across different proximity samples. We examine the\nproblem over 504 pretrained ImageNet models and observe that: 1) Proximity bias\nexists across a wide variety of model architectures and sizes; 2)\nTransformer-based models are relatively more susceptible to proximity bias than\nCNN-based models; 3) Proximity bias persists even after performing popular\ncalibration algorithms like temperature scaling; 4) Models tend to overfit more\nheavily on low proximity samples than on high proximity samples. Motivated by\nthe empirical findings, we propose ProCal, a plug-and-play algorithm with a\ntheoretical guarantee to adjust sample confidence based on proximity. To\nfurther quantify the effectiveness of calibration algorithms in mitigating\nproximity bias, we introduce proximity-informed expected calibration error\n(PIECE) with theoretical analysis. We show that ProCal is effective in\naddressing proximity bias and improving calibration on balanced, long-tail, and\ndistribution-shift settings under four metrics over various model\narchitectures. We believe our findings on proximity bias will guide the\ndevelopment of *fairer and better-calibrated* models, contributing to the\nbroader pursuit of trustworthy AI. Our code is available at:\nhttps://github.com/MiaoXiong2320/ProximityBias-Calibration.\n", " In this study, a hybrid method based on an Adaptive Neuro-Fuzzy Inference\nSystem (ANFIS) and Particle Swarm Optimization (PSO) for diagnosing Liver\ndisorders (ANFIS-PSO) is introduced. This smart diagnosis method deals with a\ncombination of making an inference system and optimization process which tries\nto tune the hyper-parameters of ANFIS based on the data-set. The Liver diseases\ncharacteristics are taken from the UCI Repository of Machine Learning\nDatabases. The number of these characteristic attributes are 7, and the sample\nnumber is 354. The right diagnosis performance of the ANFIS-PSO intelligent\nmedical system for liver disease is evaluated by using classification accuracy,\nsensitivity and specificity analysis, respectively. According to the\nexperimental results, the performance of ANFIS-PSO can be more considerable\nthan traditional FIS and ANFIS without optimization phase.\n", " Molecular property prediction is a critical task in computational drug\ndiscovery. While recent advances in Graph Neural Networks (GNNs) and\nTransformers have shown to be effective and promising, they face the following\nlimitations: Transformer self-attention does not explicitly consider the\nunderlying molecule structure while GNN feature representation alone is not\nsufficient to capture granular and hidden interactions and characteristics that\ndistinguish similar molecules. To address these limitations, we propose SYN-\nFUSION, a novel approach that synergistically combines pre-trained features\nfrom GNNs and Transformers. This approach provides a comprehensive molecular\nrepresentation, capturing both the global molecule structure and the individual\natom characteristics. Experimental results on MoleculeNet benchmarks\ndemonstrate superior performance, surpassing previous models in 5 out of 7\nclassification datasets and 4 out of 6 regression datasets. The performance of\nSYN-FUSION has been compared with other Graph-Transformer models that have been\njointly trained using a combination of transformer and graph features, and it\nis found that our approach is on par with those models in terms of performance.\nExtensive analysis of the learned fusion model across aspects such as loss,\nlatent space, and weight distribution further validates the effectiveness of\nSYN-FUSION. Finally, an ablation study unequivocally demonstrates that the\nsynergy achieved by SYN-FUSION surpasses the performance of its individual\nmodel components and their ensemble, offering a substantial improvement in\npredicting molecular properties.\n", " Let $p(z)=z^n+a_{n-1}z^{n-1}+a_{n-2}z^{n-2}+\\ldots+a_1z+a_0$ be a complex\npolynomial with $a_0\\neq 0$ and $n\\geq 3$. Several new upper bounds for the\nmoduli of the zeros of $p$ are developed. In particular, if\n$\\alpha=\\sqrt{\\sum_{j=0}^{n-1}|a_j|^2}$ and $z$ is any zero of $p$, then we\nshow that \\begin{eqnarray*}\n |z|^2 &\\leq & \\cos^2 \\frac{\\pi}{n+1}+|a_{n-2}|+ \\frac{1}{4} \\left (\n|a_{n-1}|+ { \\alpha} \\right)^2 + \\frac{1}{2}\\sqrt{\\alpha^2-|a_{n-1}|^2} +\n\\frac{1}{2}{\\alpha}, \\end{eqnarray*} which is sharper than the Abu-Omar and\nKittaneh's bound \\begin{eqnarray*}\n |z|^2 &\\leq & \\cos^2 \\frac{\\pi}{n+1}+ \\frac{1}{4} \\left ( |a_{n-1}|+ {\n\\alpha}\\right)^2 + {\\alpha} \\end{eqnarray*} if and only if $2|a_{n-2}|<\n\\sqrt{\\sum_{j=0}^{n-1}|a_j|^2}-\\sqrt{\\sum_{j=0}^{n-2}|a_j|^2}. $ The upper\nbounds obtained here enable us to describe smaller annuli in the complex plane\ncontaining all the zeros of $p$.\n", " In this paper we consider the so-called crystallographic Radon transform (or\ncrystallographic $X$-ray transform) and totally geodesic Radon transform on the\ngroup of rotations SO(3). As we show both of these transforms naturally appear\nin texture analysis, i.e. the analysis of preferred crystallographic\norientation. Although we discuss only applications to texture analysis both\ntransforms have other applications as well.\n In section 2 we start with motivations and applications. In sections 3 and 4\nwe develop a general framework on compact Lie groups. In section 5 we give a\ndetailed analysis of the totally geodesic Radon transform on SO(3). In section\n\\ref{relations} we compare crystallographic Radon transform on SO(3) and Funk\ntransform on $S^{3}$. In section \\ref{1} we show non-invertibility of the\ncrystallographic transform. In section 8 we describe an exact reconstruction\nformula for bandlimited functions, which uses only a finite number of samples\nof their Radon transform. Some auxiliary results for this section are collected\nin Appendix.\n", " A Matlab program is presented that computes derivative corrections in the\nS-dual invariant formulation for IIB graviton scattering to any order in\nperturbation theory. The coefficients of the four-point function are produced,\npertaining to the non-logarithmic terms. The program can be modified to find\ncoefficients of the higher-point functions. Instantons have not explicitly been\nincluded.\n", " In the classical linear degeneracy testing problem, we are given $n$ real\nnumbers and a $k$-variate linear polynomial $F$, for some constant $k$, and\nhave to determine whether there exist $k$ numbers $a_1,\\ldots,a_k$ from the set\nsuch that $F(a_1,\\ldots,a_k) = 0$. We consider a generalization of this problem\nin which $F$ is an arbitrary constant-degree polynomial, we are given $k$ sets\nof $n$ numbers, and have to determine whether there exist a $k$-tuple of\nnumbers, one in each set, on which $F$ vanishes. We give the first improvement\nover the na\\\"ive $O^*(n^{k-1})$ algorithm for this problem (where the\n$O^*(\\cdot)$ notation omits subpolynomial factors).\n We show that the problem can be solved in time $O^*\\left( n^{k - 2 + \\frac\n4{k+2}}\\right)$ for even $k$ and in time $O^*\\left( n^{k - 2 +\n\\frac{4k-8}{k^2-5}}\\right)$ for odd $k$ in the real RAM model of computation.\nWe also prove that for $k=4$, the problem can be solved in time\n$O^*(n^{2.625})$ in the algebraic decision tree model, and for $k=5$ it can be\nsolved in time $O^*(n^{3.56})$ in the same model, both improving on the above\nuniform bounds.\n All our results rely on an algebraic generalization of the standard\nmeet-in-the-middle algorithm for $k$-SUM, powered by recent algorithmic\nadvances in the polynomial method for semi-algebraic range searching. In fact,\nour main technical result is much more broadly applicable, as it provides a\ngeneral tool for detecting incidences and other interactions between points and\nalgebraic surfaces in any dimension. In particular, it yields an efficient\nalgorithm for a general, algebraic version of Hopcroft's point-line incidence\ndetection problem in any dimension.\n", " The phenomenon of unipolar induction consists in generating electric field by\nuniformly rotating permanent magnets made of either conducting or dielectric\nsubstance. The origin of the field around conducting magnets is explained and\ncalculated in Sect. 1. Dielectrics are considered in Sect. 2, where the role of\nthe electric moment due to moving magnetic dipole moment is clarified. Sect. 3\nis devoted to the description of the effect in the corotating frame of\nreference. In Sect. 4 the theory is applied to the Earth core and it is shown\nthat the calculated electric field falls precisely into the range of the\nambient electric field observed in the Tethered Satellite System mission of\nNASA. In the last paragraph I reflect on whether flyby anomaly might be\nattributed to an electric effect.\n", " We study the Olsen model for the peroxidase-oxidase reaction. The dynamics is\nanalyzed using a geometric decomposition based upon multiple time scales. The\nOlsen model is four-dimensional, not in a standard form required by geometric\nsingular perturbation theory and contains multiple small parameters. These\nthree obstacles are the main challenges we resolve by our analysis. Scaling and\nthe blow-up method are used to identify several subsystems. The results\npresented here provide a rigorous analysis for two oscillatory modes. In\nparticular, we prove the existence of non-classical relaxation oscillations in\ntwo cases. The analysis is based upon desingularization of lines of\ntranscritical and submanifolds of fold singularities in combination with an\nintegrable relaxation phase. In this context our analysis also explains an\nassumption that has been utilized, based purely on numerical reasoning, in a\nprevious bifurcation analysis by Desroches, Krauskopf and Osinga\n[{Discret.}{Contin.}{Dyn.}{Syst.}S, 2(4), p.807--827, 2009]. Furthermore, the\ngeometric decomposition we develop forms the basis to prove the existence of\nmixed-mode and chaotic oscillations in the Olsen model, which will be discussed\nin more detail in future work.\n", " We consider the $X(3872)$ resonance as a $J^{PC}=1^{++}$ $D\\bar D^*$ hadronic\nmolecule. According to heavy quark spin symmetry, there will exist a partner\nwith quantum numbers $2^{++}$, $X_{2}$, which would be a $D^*\\bar D^*$ loosely\nbound state. The $X_{2}$ is expected to decay dominantly into $D\\bar D$, $D\\bar\nD^*$ and $\\bar D D^*$ in $d$-wave. In this work, we calculate the decay widths\nof the $X_{2}$ resonance into the above channels, as well as those of its\nbottom partner, $X_{b2}$, the mass of which comes from assuming heavy flavor\nsymmetry for the contact terms. We find partial widths of the $X_{2}$ and\n$X_{b2}$ of the order of a few MeV. Finally, we also study the radiative\n$X_2\\to D\\bar D^{*}\\gamma$ and $X_{b2} \\to \\bar B B^{*}\\gamma$ decays. These\ndecay modes are more sensitive to the long-distance structure of the resonances\nand to the $D\\bar D^{*}$ or $B\\bar B^{*}$ final state interaction.\n", " By extending the notion of Arens regularity of bilinear mappings, we say that\na bounded trilinear map on Banach spaces id Aron--Berner regular when all its\nsix Aron-Berner extensions to the bidual spaces coincide. We give some results\non the Aron-Berner regularity of certain trilinear maps. We then focus on the\nbidual, $E^{**},$ of a Jordan banach triple system $(E,\\pi)$, and investigate\nthose conditions under which $E^{**}$ is itself a Jordan Banach triple system\nunder each of the Aron-Berner extensions of the triple product $\\pi.$ We also\ncompare these six triple products with those arising from certain ultrafilters\nbased on the ultrapower formulation of the principle of local reflexivity. In\nparticular, we examine the Aron--Berner triple products on the bidual of a\nJB$^*$-triple in relation with the so-called Dineen's theorem. Some\nilluminating examples are included and some questions are also left undecided.\n", " We give a brief review of direct $CP$ violation effects in charmless $B$\ndecays, both at the inclusive and at the exclusive level. It is emphasized that\ntypical exclusive final states are composed of 2--4 $K$ or $\\pi$'s, hence\nparticle identification (in terms of $K/\\pi$ separation) is of crucial\nimportance for studying $CP$ violation in these modes. Another kind of\n``charmless\" final state is also emphasized: $b\\to dc\\bar c$. The decay rate is\ntree dominant, while ``$q^2$\" is necessarily above $c\\bar c$ threshold hence in\nthe domain of perturbative QCD. The $CP$ asymmetry is therefore relatively\nclean and reliable. Theoretically, a few $\\times 10^7$ $B$'s would suffice if\none could accumulate modes like $B^+\\to D^{+\\left(*\\right)} \\bar\nD^{0\\left(*\\right)}$. [To appear in proceedings of ``The Third KEK Topical\nConference on\n CP Violation --- Its Implications to Particle Physics and Cosmology'',\n 16--18 November, 1993, KEK, Tsukuba, Japan (Eds. Y. Kuno and Y. Okada).]\n", " The Turing test for comparing computer performance to that of humans is well\nknown, but, surprisingly, there is no widely used test for comparing how much\nbetter human-computer systems perform relative to humans alone, computers\nalone, or other baselines. Here, we show how to perform such a test using the\nratio of means as a measure of effect size. Then we demonstrate the use of this\ntest in three ways. First, in an analysis of 79 recently published experimental\nresults, we find that, surprisingly, over half of the studies find a decrease\nin performance, the mean and median ratios of performance improvement are both\napproximately 1 (corresponding to no improvement at all), and the maximum ratio\nis 1.36 (a 36% improvement). Second, we experimentally investigate whether a\nhigher performance improvement ratio is obtained when 100 human programmers\ngenerate software using GPT-3, a massive, state-of-the-art AI system. In this\ncase, we find a speed improvement ratio of 1.27 (a 27% improvement). Finally,\nwe find that 50 human non-programmers using GPT-3 can perform the task about as\nwell as--and less expensively than--the human programmers. In this case,\nneither the non-programmers nor the computer would have been able to perform\nthe task alone, so this is an example of a very strong form of human-computer\nsynergy.\n", " Solar activity It is vital to accurately predict solar activity, in order to\ndecrease the plausible damage of electronic equipment in the event of a large\nhigh-intensity solar eruption. Recently, we have proposed BELFIS (Brain\nEmotional Learning-based Fuzzy Inference System) as a tool for the forecasting\nof chaotic systems. The structure of BELFIS is designed based on the neural\nstructure of fear conditioning. The function of BELFIS is implemented by\nassigning adaptive networks to the components of the BELFIS structure. This\npaper especially focuses on performance evaluation of BELFIS as a predictor by\nforecasting solar cycles 16 to 24. The performance of BELFIS is compared with\nother computational models used for this purpose, and in particular with\nadaptive neuro-fuzzy inference system (ANFIS).\n", " This paper introduces a novel operator, termed the Y operator, to elevate\ncontrol performance in Actor-Critic(AC) based reinforcement learning for\nsystems governed by stochastic differential equations(SDEs). The Y operator\ningeniously integrates the stochasticity of a class of child-mother system into\nthe Critic network's loss function, yielding substantial advancements in the\ncontrol performance of RL algorithms.Additionally, the Y operator elegantly\nreformulates the challenge of solving partial differential equations for the\nstate-value function into a parallel problem for the drift and diffusion\nfunctions within the system's SDEs.A rigorous mathematical proof confirms the\noperator's validity.This transformation enables the Y Operator-based\nReinforcement Learning(YORL) framework to efficiently tackle optimal control\nproblems in both model-based and data-driven systems.The superiority of YORL is\ndemonstrated through linear and nonlinear numerical examples showing its\nenhanced performance over existing methods post convergence.\n", " A strong edge coloring of a graph is a proper edge coloring in which every\ncolor class is an induced matching. The strong chromatic index $\\chi_s'(G)$ of\na graph $G$ is the minimum number of colors in a strong edge coloring of $G$.\nLet $\\Delta \\geq 4$ be an integer. In this note, we study the odd graphs and\nshow the existence of some special walks. By using these results and Chang's\nideas in [Discuss. Math. Graph Theory 34 (4) (2014) 723--733], we show that\nevery planar graph with maximum degree at most $\\Delta$ and girth at least $10\n\\Delta - 4$ has a strong edge coloring with $2\\Delta - 1$ colors. In addition,\nwe prove that if $G$ is a graph with girth at least $2\\Delta - 1$ and mad$(G) <\n2 + \\frac{1}{3\\Delta - 2}$, where $\\Delta$ is the maximum degree and $\\Delta\n\\geq 4$, then $\\chi_s'(G) \\leq 2\\Delta - 1$, if $G$ is a subcubic graph with\ngirth at least $8$ and mad$(G) < 2 + \\frac{2}{23}$, then $\\chi_s'(G) \\leq 5$.\n", " One of the main indications for New Physics in rare $B$-decays is deduced\nfrom the tension between experimental and Standard Model predictions of the\nangular analysis of the $B^0 \\to K^{*0} \\mu^+\\mu^-$ decay. There are however\npossible non-local hadronic effects which in principle can also explain these\ntensions. In this work, we consider a statistical approach for differentiating\nthe source of the tension in $B^0 \\to K^{*0} \\mu^+\\mu^-$ observables and we\nalso investigate the prospects of such a comparison with future data from the\nLHCb experiment.\n", " We present a complete analysis of Lepton Flavor Universality (LFU) violations\nin leptonic $\\tau$ decays in motivated models addressing the $B$-physics\nanomalies, based on the $SU(4)\\times SU(3)\\times SU(2)\\times U(1)$ gauge group.\nWe show that the inclusion of vector-like fermions, required by $B$-physics\ndata, leads to sizable modifications of the leading-log results derived within\nan Effective Field Theory approach. In the motivated parameter-space region\nrelevant to the $B$-physics anomalies, the models predict a few per-mil\ndecrease of the effective $W$-boson coupling to $\\tau$, within the reach of\nfuture experiments.\n", " We pose the following conjecture:\n (*) If A is the union of line segments in R^n, and B is the union of the\ncorresponding full lines then the Hausdorff dimensions of A and B agree.\n We prove that this conjecture would imply that every Besicovitch set (compact\nset that contains line segments in every direction) in R^n has Hausdorff\ndimension at least n-1 and (upper) Minkowski dimension n. We also prove that\nconjecture (*) holds if the Hausdorff dimension of B is at most 2, so in\nparticular it holds in the plane.\n", " Traffic signal control (TSC) is a challenging problem within intelligent\ntransportation systems and has been tackled using multi-agent reinforcement\nlearning (MARL). While centralized approaches are often infeasible for\nlarge-scale TSC problems, decentralized approaches provide scalability but\nintroduce new challenges, such as partial observability. Communication plays a\ncritical role in decentralized MARL, as agents must learn to exchange\ninformation using messages to better understand the system and achieve\neffective coordination. Deep MARL has been used to enable inter-agent\ncommunication by learning communication protocols in a differentiable manner.\nHowever, many deep MARL communication frameworks proposed for TSC allow agents\nto communicate with all other agents at all times, which can add to the\nexisting noise in the system and degrade overall performance. In this study, we\npropose a communication-based MARL framework for large-scale TSC. Our framework\nallows each agent to learn a communication policy that dictates \"which\" part of\nthe message is sent \"to whom\". In essence, our framework enables agents to\nselectively choose the recipients of their messages and exchange variable\nlength messages with them. This results in a decentralized and flexible\ncommunication mechanism in which agents can effectively use the communication\nchannel only when necessary. We designed two networks, a synthetic $4 \\times 4$\ngrid network and a real-world network based on the Pasubio neighborhood in\nBologna. Our framework achieved the lowest network congestion compared to\nrelated methods, with agents utilizing $\\sim 47-65 \\%$ of the communication\nchannel. Ablation studies further demonstrated the effectiveness of the\ncommunication policies learned within our framework.\n", " Motivation: Usefulness of analysis derived from Affymetrix microarrays\ndepends largely upon the reliability of files describing the correspondence\nbetween probe sets, genes and transcripts. In particular, in case a gene is\ntargeted by two probe sets, one must be able to assess if the corresponding\nsignals measure a group of common transcripts or two groups of transcripts with\nlittle or no overlap.\n Results: Probe sets that effectively target the same group of transcripts\nhave specific properties in the trancriptional networks we constructed. We\nfound indeed that such probe sets had a very low negative correlation, a high\npositive correlation and a similar neighbourhood. Taking advantage of these\nproperties, we devised a test allowing to group probe sets which target the\nsame group of transcripts in a particular network. By considering several\nnetworks, additional information concerning the frequency of these associations\nwas obtained.\n Availability and Implementation: The programs developed in Python (PSAWNpy)\nand in Matlab (PSAWNml) are freely available, and can be downloaded at\nhttp://code.google.com/p/arraymatic/.\n Tutorials and reference manuals are available at\nhttp://bns.crbm.cnrs.fr/softwares.html.\n Contact: [email protected].\n Supplementary information: Supplementary data are available at\nhttp://bns.crbm.cnrs.fr/download.html.\n", " We consider the problem of continuous quantum measurement of coherent\noscillations between two quantum states of an individual two-state system. It\nis shown that the interplay between the information acquisition and the\nbackaction dephasing of the oscillations by the detector imposes a fundamental\nlimit, equal to 4, on the signal-to-noise ratio of the measurement. The limit\nis universal, e.g., independent of the coupling strength between the detector\nand system, and results from the tendency of quantum measurement to localize\nthe system in one of the measured eigenstates.\n", " A dynamical system obtains a wide variety of kinetic realizations, which is\nadvantageous for the analysis of biochemical systems. A reaction network,\nderived from a dynamical system, may or may not possess some properties needed\nfor a thorough analysis. We improve and extend the work of M. Johnston\n\\cite{JOHN2014} and Hong et al. \\cite{HONG2023} on network translations to\nnetwork transformations, where the network is modified while preserving the\ndynamical system. These transformations can shrink, extend, or retain the\nstoichiometric subspace. Here, we show that positive dependent network can be\ntranslated to a weakly reversible network. Using the kinetic realizations of\n(1) calcium signaling in the olfactory system and (2) metabolic insulin\nsignaling, we demonstrate the benefits of transformed systems with positive\ndeficiency for analyzing biochemical systems. Furthermore, we present an\nalgorithm for a network transformation of a weakly reversible non-complex\nfactorizable kinetic (NFK) system to a weakly reversible complex factorizable\nkinetic (CFK) system, thereby enhancing the Subspace Coincidence Theorem for\nNFK systems of Nazareno et al. \\cite{NAZA2019}. Finally, using the transformed\nkinetic realization of monolignol biosynthesis in \\textit{Populus xylem}, we\nstudy the structural and kinetic properties of transformed systems,including\nthe invariance of concordance and variation of injectivity and\nmono-/multi-stationarity under network transformation.\n", " This paper is a short overview of the main Abelian- and Tauberian-type\nresults from [4, 14, 26] regarding the asymptotic analysis of different classes\nof generalized functions in terms of appropriate frames. The Tauberian-type\nresults provide a comprehensive characterization of the quasiasymptotic and\nS-asymptotic properties of distributions.\n", " Using the same quenched limit as input, we compare the charmonium spectra\npredicted by the coupled-channel model and the screened potential model in the\nmass region below 4 GeV, in which the contributions from decay channels\ninvolving P-wave (as well as even higher excited) D mesons can be neglected. We\nfind that the two models have similar global features in describing charmonium\nspectrum since they embody the same effect of the vacuum polarization of\ndynamical light quark pairs. This agreement is encouraging for using the\nscreened potential to further describe the mass spectrum of higher charmonium\nin the region above 4 GeV. Applying both models will be helpful to clarify the\nnature of the newly discovered \"X,Y,Z\" mesons in B-factories and other\nexperiments. In particular, we show the S-wave decay coupling effect on\nlowering the $\\chi_{c1}(2P)$ mass towards the $D\\bar {D^*}$ threshold, in\nsupport of the assignment of the X(3872) as a $\\chi_{c1}(2P)$-dominated\ncharmonium state.\n", " We find a two-parameter family of coupled Painlev\\'e systems in dimension\nfour with affine Weyl group symmetry of type $A_4^{(2)}$. For a degenerate\nsystem of $A_4^{(2)}$ system, we also find a one-parameter family of coupled\nPainlev\\'e systems in dimension four with affine Weyl group symmetry of type\n$A_1^{(1)}$. We show that for each system, we give its symmetry and holomorphy\nconditions. These symmetries, holomorphy conditions and invariant divisors are\nnew. Moreover, we find a one-parameter family of partial differential systems\nin three variables with $W(A_1^{(1)})$-symmetry. We show the relation between\nits polynomial Hamiltonian system and an autonomous version of the system of\ntype $A_1^{(1)}$.\n", " Storage architectures ranging from minimum bandwidth regenerating encoded\ndistributed storage systems to declustered-parity RAIDs can be designed using\ndense partial Steiner systems in order to support fast reads, writes, and\nrecovery of failed storage units. In order to ensure good performance,\npopularities of the data items should be taken into account and the frequencies\nof accesses to the storage units made as uniform as possible. A proposed\ncombinatorial model ranks items by popularity and assigns data items to\nelements in a dense partial Steiner system so that the sums of ranks of the\nelements in each block are as equal as possible. By developing necessary\nconditions in terms of independent sets, we demonstrate that certain Steiner\nsystems must have a much larger difference between the largest and smallest\nblock sums than is dictated by an elementary lower bound. In contrast, we also\nshow that certain dense partial $S(t, t+1, v)$ designs can be labeled to\nrealize the elementary lower bound. Furthermore, we prove that for every\nadmissible order $v$, there is a Steiner triple system $(S(2, 3, v))$ whose\nlargest difference in block sums is within an additive constant of the lower\nbound.\n", " Recent advancements in natural language processing have led to the\nproliferation of large language models (LLMs). These models have been shown to\nyield good performance, using in-context learning, even on tasks and languages\nthey are not trained on. However, their performance on African languages is\nlargely understudied relative to high-resource languages. We present an\nanalysis of four popular large language models (mT0, Aya, LLaMa 2, and GPT-4)\non six tasks (topic classification, sentiment classification, machine\ntranslation, summarization, question answering, and named entity recognition)\nacross 60 African languages, spanning different language families and\ngeographical regions. Our results suggest that all LLMs produce lower\nperformance for African languages, and there is a large gap in performance\ncompared to high-resource languages (such as English) for most tasks. We find\nthat GPT-4 has an average to good performance on classification tasks, yet its\nperformance on generative tasks such as machine translation and summarization\nis significantly lacking. Surprisingly, we find that mT0 had the best overall\nperformance for cross-lingual QA, better than the state-of-the-art supervised\nmodel (i.e. fine-tuned mT5) and GPT-4 on African languages. Similarly, we find\nthe recent Aya model to have comparable result to mT0 in almost all tasks\nexcept for topic classification where it outperform mT0. Overall, LLaMa 2\nshowed the worst performance, which we believe is due to its English and\ncode-centric~(around 98%) pre-training corpus. Our findings confirms that\nperformance on African languages continues to remain a hurdle for the current\nLLMs, underscoring the need for additional efforts to close this gap.\n", " Recently, there has been interest in the question of whether a partial matrix\nin which many of the fully defined principal submatrices are PSD is\napproximately PSD completable. These questions are related to graph theory\nbecause we can think of the entries of a symmetric matrix as corresponding to\nthe edges of a graph.\n We first introduce a family of graphs, which we call thickened graphs; these\ncontain both triangle-free and chordal graphs, and can be viewed as the result\nof replacing the edges of a graph by an arbitrary chordal graph. We believe\nthese graphs might be of independent interest.\n We then show that for a class of graphs including thickened graphs, it is\npossible to get quantitative bounds on how well the property of having these\nprincipal submatrices being PSD approximates the PSD-completability property.\nThese bounds frequently only depend on the size of the smallest cycle of size\nat least 4 in the graph. We introduce some tools that allow us to better\ncontrol the quality of these approximations and indicate how these\napproximations can be used to improve the performance of semidefinite programs.\nThe tools we use in this paper are an interesting mix of algebraic topology,\nstructural graph theory, and spectral analysis.\n", " The present article is a review of recent developments concerning the notion\nof F{\\o}lner sequences both in operator theory and operator algebras. We also\ngive a new direct proof that any essentially normal operator has an increasing\nF{\\o}lner sequence $\\{P_n\\}$ of non-zero finite rank projections that strongly\nconverges to 1. The proof is based on Brown-Douglas-Fillmore theory. We use\nF{\\o}lner sequences to analyze the class of finite operators introduced by\nWilliams in 1970. In the second part of this article we examine a procedure of\napproximating any amenable trace on a unital and separable C*-algebra by\ntracial states $\\mathrm{Tr}(\\cdot P_n)/\\mathrm{Tr}(P_n)$ corresponding to a\nF{\\o}lner sequence and apply this method to improve spectral approximation\nresults due to Arveson and B\\'edos. The article concludes with the analysis of\nC*-algebras admitting a non-degenerate representation which has a F{\\o}lner\nsequence or, equivalently, an amenable trace. We give an abstract\ncharacterization of these algebras in terms of unital completely positive maps\nand define F{\\o}lner C*-algebras as those unital separable C*-algebras that\nsatisfy these equivalent conditions. This is analogous to Voiculescu's abstract\ncharacterization of quasidiagonal C*-algebras.\n", " The multiple futile cycle is a phosphorylation system in which a molecular\nsubstrate might be phosphorylated sequentially n times by means of an enzymatic\nmechanism. The system has been studied mathematically using reaction network\ntheory and ordinary differential equations. It is known that the system might\nhave at least as many as 2[n/2]+1 steady states (where [x] is the integer part\nof x) for particular choices of parameters. Furthermore, for the simple and\ndual futile cycles (n=1,2) the stability of the steady states has been\ndetermined in the sense that the only steady state of the simple futile cycle\nis globally stable, while there exist parameter values for which the dual\nfutile cycle admits two asymptotically stable and one unstable steady state.\nFor general n, evidence that the possible number of asymptotically stable\nsteady states increases with $n$ has been given, which has led to the\nconjecture that parameter values can be chosen such that [n/2]+1 out of\n2[n/2]+1 steady states are asymptotically stable and the remaining steady\nstates are unstable.\n We prove this conjecture here by first reducing the system to a smaller one,\nfor which we find a choice of parameter values that give rise to a unique\nsteady state with multiplicity 2[n/2]+1. Using arguments from geometric\nsingular perturbation theory, and a detailed analysis of the centre manifold of\nthis steady state, we achieve the desired result.\n", " We study the asymptotic clustering (phase-locking) dynamics for the Kuramoto\nmodel. For the analysis of emergent asymptotic patterns in the Kuramoto flow,\nwe introduce the pathwise critical coupling strength which yields a sharp\ntransition from partial phase-locking to complete phase-locking, and provide\nnontrivial upper bounds for the pathwise critical coupling strength. Numerical\nsimulations suggest that multi- and mono-clusters can emerge asymptotically in\nthe Kuramoto flow depending on the relative magnitude of the coupling strength\ncompared to the sizes of natural frequencies. However, theoretical and rigorous\nanalysis for such phase-locking dynamics of the Kuramoto flow still lacks a\ncomplete understanding, although there were some recent progress on the\ncomplete synchronization of the Kuramoto model in a sufficiently large coupling\nstrength regime. In this paper, we present sufficient frameworks for partial\nphase-locking of a majority ensemble and the complete phase-locking in terms of\nthe initial phase configuration, coupling strength and natural frequencies. As\na by-product of our analysis, we obtain nontrivial upper bounds for the\npathwise critical coupling strength in terms of the diameter of natural\nfrequencies, initial Kuramoto order parameter and the system size $N$. We also\nshow that phase-locked states whose order parameters are less than\n$N^{-\\frac{1}{2}}$ are linearly unstable.\n", " Diffusive transport is among the most common phenomena in nature [1].\nHowever, as predicted by Anderson [2], diffusion may break down due to\ninterference. This transition from diffusive transport to localization of waves\nshould occur for any type of classical or quantum wave in any media as long as\nthe wavelength becomes comparable to the transport mean free path $\\ell^*$ [3].\nThe signatures of localization and those of absorption, or bound states, can\nhowever be similar, such that an unequivocal proof of the existence of wave\nlocalization in disordered bulk materials is still lacking. Here we present\nmeasurements of time resolved non-classical diffusion of visible light in\nstrongly scattering samples, which cannot be explained by absorption, sample\ngeometry or reduction in transport velocity. Deviations from classical\ndiffusion increase strongly with decreasing $\\ell^*$ as expected for a phase\ntransition. This constitutes an experimental realization of the critical regime\nin the approach to Anderson localization.\n", " In this work we give a review of the original formulation of the relativistic\nwave equation for particles with spin one-half. Traditionally \\`a la Dirac,\nit's proposed that the ``square root'' of the Klein-Gordon (K-G) equation\ninvolves a 4 component (Dirac) spinor and in the non-relativistic limit it can\nbe written as 2 equations for two 2 component spinors. On the other hand, there\nexists Weyl's formalism, in which one works from the beginning with 2 component\nWeyl spinors, which are the fundamental objects of the helicity formalism. In\nthis work we rederive Weyl's equations directly, starting from K-G equation. We\nalso obtain the electromagnetic interaction through minimal coupling and we get\nthe interaction with the magnetic moment. As an example of the use of that\nformalism, we calculate Compton scattering using the helicity methods.\n", " For the analysis of systems consisting of small, regular objects, the methods\nof mathematical morphology applied to images of these systems are well-suited.\nOne of these methods is the use of Voronoi polygons. It was found that the\nVoronoi tessellation method represents a powerful tool for the analysis of thin\nfilm morphology and provides nanostructural information to many multi-particle\nassemblies. In these notes, several morphological algorithms are analyzed and\nwe study how to join all of them to design a graphical user interface (GUI)\nthat provides as input for the system the \"AFM image\" and interprets the output\nof the system in terms of errors and generators coordinates.\n", " In this paper I have given a mathematical model of Cell reprogramming from a\ndifferent contexts. Here I considered there is a delay in differential\nregulator rate equations due to intermediate regulator's regulations. At first\nI gave some basic mathematical models by Ferell Jr.[2] of reprogramming and\nafter that I gave mathematical model of cell reprogramming by Mithun Mitra[4].\nIn the last section I contributed a mathematical model of cell reprogramming\nfrom intermediate steps regulations and tried to find the critical point of\npluripotent cell.\n", " The case difference heuristic (CDH) approach is a knowledge-light method for\nlearning case adaptation knowledge from the case base of a case-based reasoning\nsystem. Given a pair of cases, the CDH approach attributes the difference in\ntheir solutions to the difference in the problems they solve, and generates\nadaptation rules to adjust solutions accordingly when a retrieved case and new\nquery have similar problem differences. As an alternative to learning\nadaptation rules, several researchers have applied neural networks to learn to\npredict solution differences from problem differences. Previous work on such\napproaches has assumed that the feature set describing problems is predefined.\nThis paper investigates a two-phase process combining deep learning for feature\nextraction and neural network based adaptation learning from extracted\nfeatures. Its performance is demonstrated in a regression task on an image\ndata: predicting age given the image of a face. Results show that the combined\nprocess can successfully learn adaptation knowledge applicable to nonsymbolic\ndifferences in cases. The CBR system achieves slightly lower performance\noverall than a baseline deep network regressor, but better performance than the\nbaseline on novel queries.\n", " In recent years, several algorithms for system identification with neural\nstate-space models have been introduced. Most of the proposed approaches are\naimed at reducing the computational complexity of the learning problem, by\nsplitting the optimization over short sub-sequences extracted from a longer\ntraining dataset. Different sequences are then processed simultaneously within\na minibatch, taking advantage of modern parallel hardware for deep learning. An\nissue arising in these methods is the need to assign an initial state for each\nof the sub-sequences, which is required to run simulations and thus to evaluate\nthe fitting loss. In this paper, we provide insights for calibration of neural\nstate-space training algorithms based on extensive experimentation and analyses\nperformed on two recognized system identification benchmarks. Particular focus\nis given to the choice and the role of the initial state estimation. We\ndemonstrate that advanced initial state estimation techniques are really\nrequired to achieve high performance on certain classes of dynamical systems,\nwhile for asymptotically stable ones basic procedures such as zero or random\ninitialization already yield competitive performance.\n", " This work introduces a benchmark assessing the performance of clustering\nGerman text embeddings in different domains. This benchmark is driven by the\nincreasing use of clustering neural text embeddings in tasks that require the\ngrouping of texts (such as topic modeling) and the need for German resources in\nexisting benchmarks. We provide an initial analysis for a range of pre-trained\nmono- and multilingual models evaluated on the outcome of different clustering\nalgorithms. Results include strong performing mono- and multilingual models.\nReducing the dimensions of embeddings can further improve clustering.\nAdditionally, we conduct experiments with continued pre-training for German\nBERT models to estimate the benefits of this additional training. Our\nexperiments suggest that significant performance improvements are possible for\nshort text. All code and datasets are publicly available.\n", " We prove that the order of the double reduced 2-factor transfer digraph\n${\\cal R}^{**}_{m}$ which is needed for the enumeration of the spanning unions\nof cycles in the rectangular grid graph $P_m \\times P_n$ ($m,n \\in N$), when\n$m$ is odd, is equal to $\\displaystyle \\mid V({\\cal R}^{**}_{m}) \\mid =\n\\frac{1}{2} \\left[{m+1 \\choose (m-1)/2 } + {(m+1)/2 \\choose \\lfloor (m+1)/4\n\\rfloor}\\right].$\n", " We present an extensive analysis of a model in which the (Majorana) Dark\nMatter candidate is a mixture between a SU(2) singlet and two SU(2) doublets.\nThis kind of setup takes the name of singlet-doublet model. We will investigate\nin detail an extension of this model in which the Dark Matter sector\ninteractions with a 2-doublet Higgs sector enforcing the complementarity\nbetween Dark Matter phenomenology and searches of extra Higgs bosons.\n", " Interaction between proteins is a fundamental mechanism that underlies\nvirtually all biological processes. Many important interactions are conserved\nacross a large variety of species. The need to maintain interaction leads to a\nhigh degree of co-evolution between residues in the interface between partner\nproteins. The inference of protein-protein interaction networks from the\nrapidly growing sequence databases is one of the most formidable tasks in\nsystems biology today. We propose here a novel approach based on the\nDirect-Coupling Analysis of the co-evolution between inter-protein residue\npairs. We use ribosomal and trp operon proteins as test cases: For the small\nresp. large ribosomal subunit our approach predicts protein-interaction\npartners at a true-positive rate of 70% resp. 90% within the first 10\npredictions, with areas of 0.69 resp. 0.81 under the ROC curves for all\npredictions. In the trp operon, it assigns the two largest interaction scores\nto the only two interactions experimentally known. On the level of residue\ninteractions we show that for both the small and the large ribosomal subunit\nour approach predicts interacting residues in the system with a true positive\nrate of 60% and 85% in the first 20 predictions. We use artificial data to show\nthat the performance of our approach depends crucially on the size of the joint\nmultiple sequence alignments and analyze how many sequences would be necessary\nfor a perfect prediction if the sequences were sampled from the same model that\nwe use for prediction. Given the performance of our approach on the test data\nwe speculate that it can be used to detect new interactions, especially in the\nlight of the rapid growth of available sequence data.\n", " Most practitioners use a variant of the Alpha-Beta algorithm, a simple\ndepth-first pro- cedure, for searching minimax trees. SSS*, with its best-first\nsearch strategy, reportedly offers the potential for more efficient search.\nHowever, the complex formulation of the al- gorithm and its alleged excessive\nmemory requirements preclude its use in practice. For two decades, the search\nefficiency of \"smart\" best-first SSS* has cast doubt on the effectiveness of\n\"dumb\" depth-first Alpha-Beta. This paper presents a simple framework for\ncalling Alpha-Beta that allows us to create a variety of algorithms, including\nSSS* and DUAL*. In effect, we formulate a best-first algorithm using\ndepth-first search. Expressed in this framework SSS* is just a special case of\nAlpha-Beta, solving all of the perceived drawbacks of the algorithm. In\npractice, Alpha-Beta variants typically evaluate less nodes than SSS*. A new\ninstance of this framework, MTD(f), out-performs SSS* and NegaScout, the\nAlpha-Beta variant of choice by practitioners.\n", " The theory of graphons is ultimately connected with the so-called cut norm.\nIn this paper, we approach the cut norm topology via the weak* topology (when\nconsidering a predual of $L^{1}$-functions). We prove that a sequence\n$W_1,W_2,W_3,\\ldots$ of graphons converges in the cut distance if and only if\nwe have equality of the sets of weak* accumulation points and of weak* limit\npoints of all sequences of graphons $W_1',W_2',W_3',\\ldots$ that are weakly\nisomorphic to $W_1,W_2,W_3,\\ldots$. We further give a short descriptive set\ntheoretic argument that each sequence of graphons contains a subsequence with\nthe property above. This in particular provides an alternative proof of the\ntheorem of Lov\\'asz and Szegedy about compactness of the space of graphons. We\nconnect these results to \"multiway cut\" characterization of cut distance\nconvergence from [Ann. of Math. (2) 176 (2012), no. 1, 151-219].\n These results are more naturally phrased in the Vietoris hyperspace $K$ over\ngraphons with the weak* topology. We show that graphons with the cut distance\ntopology are homeomorphic to a closed subset of $K$, and deduce several\nconsequences of this fact.\n From these concepts a new order on the space of graphons emerges. This order\nallows to compare how structured two graphons are. We establish basic\nproperties of this \"structurdness order\".\n", " The k-means method is one of the most widely used clustering algorithms,\ndrawing its popularity from its speed in practice. Recently, however, it was\nshown to have exponential worst-case running time. In order to close the gap\nbetween practical performance and theoretical analysis, the k-means method has\nbeen studied in the model of smoothed analysis. But even the smoothed analyses\nso far are unsatisfactory as the bounds are still super-polynomial in the\nnumber n of data points.\n In this paper, we settle the smoothed running time of the k-means method. We\nshow that the smoothed number of iterations is bounded by a polynomial in n and\n1/\\sigma, where \\sigma is the standard deviation of the Gaussian perturbations.\nThis means that if an arbitrary input data set is randomly perturbed, then the\nk-means method will run in expected polynomial time on that input set.\n", " We develop a self-consistent approach to study the spectral properties of a\nclass of quantum mechanical operators by using the knowledge about monodromies\nof $2\\times 2$ linear systems (Riemann-Hilbert correspondence). Our technique\napplies to a variety of problems, though in this paper we only analyse in\ndetail two examples. First we review the case of the (modified) Mathieu\noperator, which corresponds to a certain linear system on the sphere and makes\ncontact with the Painlev\\'e $\\mathrm{III}_3$ equation. Then we extend the\nanalysis to the 2-particle elliptic Calogero-Moser operator, which corresponds\nto a linear system on the torus. By using the Kiev formula for the\nisomonodromic tau functions, we obtain the spectrum of such operators in terms\nof self-dual Nekrasov functions ($\\epsilon_1+\\epsilon_2=0$). Through blowup\nrelations, we also find Nekrasov-Shatashvili type of quantizations\n($\\epsilon_2=0$).\n In the case of the torus with one regular singularity we obtain certain\nresults which are interesting by themselves. Namely, we derive blowup equations\n(filling some gaps in the literature) and we relate them to the bilinear form\nof the isomonodromic deformation equations. In addition, we extract the\n$\\epsilon_2\\to 0$ limit of the blowup relations from the regularized action\nfunctional and CFT arguments.\n", " We reexamine lepton mixing in gauge models by considering two theories within\nthe type I seesaw mechanism, the Extended Standard Model, i.e. SU(2)_L \\times\nU(1)_Y with singlet right-handed heavy neutrinos, and the Left-Right Model\nSU(2)_L x SU(2)_R x U(1)_{B-L}. The former is often used as a simple heuristic\napproach to masses and mixing of light neutrinos and to leptogenesis, while we\nconsider the latter as an introduction to other left-right symmetric gauge\ntheories like SO(10). We compare lepton mixing in both theories for general\nparameter space and discuss also some particular cases. In the electroweak\nbroken phase, we study in parallel both models in the \"current basis\" (diagonal\ngauge interactions), and in the \"mass basis\" (diagonal mass matrices and mixing\nin the interaction), and perform the counting of CP conserving and CP violating\nparameters in both bases. We extend the analysis to the Pati-Salam model\nSU(4)_C x SU(2)_L x SU(2)_R and to SO(10). Although specifying the Higgs sector\nincreases the predictive power, in the most general case one has the same\nparameter structure in the lepton sector for all the left-right symmetric gauge\nmodels. We make explicit the differences between the Extended Standard Model\nand the left-right models, in particular CP violating and lepton-number\nviolating new terms involving the W_R gauge bosons. As expected, at low energy,\nthe differences in the light neutrino spectrum and mixing appear only beyond\nleading order in the ratio of Dirac mass to right-handed Majorana mass.\n", " In Baraka's paper [2], he obtained the Littlewood-Paley characterization of\nCampanato spaces $L^{2,\\lambda}$ and introduced $\\mathcal {L}^{p,\\lambda,s}$\nspaces. He showed that $\\mathcal\n{L}^{2,\\lambda,s}=(-\\triangle)^{-\\frac{s}{2}}L^{2,\\lambda}$ for\n$0\\leq\\lambda<n+2$. In [7], by using the properties of fractional Carleson\nmeasures, J Xiao proved that for $n\\geq2$, $0<\\alpha<1$.\n$(-\\triangle)^{-\\frac{\\alpha}{2}}L^{2,n-2\\alpha}$ is essential the\n$Q_{\\alpha}(\\mathbb{R}^n)$ spaces which were introduced in [4]. Then we could\nconclude that $Q_{\\alpha}(\\mathbb{R}^n)=\\mathcal {L}^{2,n-2\\alpha,\\alpha}$ for\n$0<\\alpha<1$. In fact, this result could be also obtained directly by using the\nmethod in [2]. In this paper, We proved this result in the spirit of [2]. This\npaper could be considered as the supplement of Baraka's work [2].\n" ]

[ " Convolutional Neural Networks have been used in a variety of image related\napplications after their rise in popularity due to ImageNet competition.\nConvolutional Neural Networks have shown remarkable results in applications\nincluding face recognition, moving target detection and tracking,\nclassification of food based on the calorie content and many more. Designing of\nConvolutional Neural Networks requires experts having a cross domain knowledge\nand it is laborious, which requires a lot of time for testing different values\nfor different hyperparameter along with the consideration of different\nconfigurations of existing architectures. Neural Architecture Search is an\nautomated way of generating Neural Network architectures which saves\nresearchers from all the brute-force testing trouble, but with the drawback of\nconsuming a lot of computational resources for a prolonged period. In this\npaper, we propose an automated Neural Architecture Search framework DQNAS,\nguided by the principles of Reinforcement Learning along with One-shot Training\nwhich aims to generate neural network architectures that show superior\nperformance and have minimum scalability problem.\n", " This paper presents a comparative study of six soft computing models namely\nmultilayer perceptron networks, Elman recurrent neural network, radial basis\nfunction network, Hopfield model, fuzzy inference system and hybrid fuzzy\nneural network for the hourly electricity demand forecast of Czech Republic.\nThe soft computing models were trained and tested using the actual hourly load\ndata for seven years. A comparison of the proposed techniques is presented for\npredicting 2 day ahead demands for electricity. Simulation results indicate\nthat hybrid fuzzy neural network and radial basis function networks are the\nbest candidates for the analysis and forecasting of electricity demand.\n", " Let F:R^n -> R be a feedforward ReLU neural network. It is well-known that\nfor any choice of parameters, F is continuous and piecewise (affine) linear. We\nlay some foundations for a systematic investigation of how the architecture of\nF impacts the geometry and topology of its possible decision regions for binary\nclassification tasks. Following the classical progression for smooth functions\nin differential topology, we first define the notion of a generic, transversal\nReLU neural network and show that almost all ReLU networks are generic and\ntransversal. We then define a partially-oriented linear 1-complex in the domain\nof F and identify properties of this complex that yield an obstruction to the\nexistence of bounded connected components of a decision region. We use this\nobstruction to prove that a decision region of a generic, transversal ReLU\nnetwork F: R^n -> R with a single hidden layer of dimension (n + 1) can have no\nmore than one bounded connected component.\n", " This paper is concerned with the utilization of deterministically modeled\nchemical reaction networks for the implementation of (feed-forward) neural\nnetworks. We develop a general mathematical framework and prove that the\nordinary differential equations (ODEs) associated with certain reaction network\nimplementations of neural networks have desirable properties including (i)\nexistence of unique positive fixed points that are smooth in the parameters of\nthe model (necessary for gradient descent), and (ii) fast convergence to the\nfixed point regardless of initial condition (necessary for efficient\nimplementation). We do so by first making a connection between neural networks\nand fixed points for systems of ODEs, and then by constructing reaction\nnetworks with the correct associated set of ODEs. We demonstrate the theory by\nconstructing a reaction network that implements a neural network with a\nsmoothed ReLU activation function, though we also demonstrate how to generalize\nthe construction to allow for other activation functions (each with the\ndesirable properties listed previously). As there are multiple types of\n\"networks\" utilized in this paper, we also give a careful introduction to both\nreaction networks and neural networks, in order to disambiguate the overlapping\nvocabulary in the two settings and to clearly highlight the role of each\nnetwork's properties.\n", " The vulnerability of networks to targeted attacks is an issue of widespread\ninterest for policymakers, military strategists, network engineers and systems\nbiologists alike. Current approaches to circumvent targeted attacks seek to\nincrease the robustness of a network by changing the network structure in one\nway or the other, leading to a higher size of the largest connected component\nfor a given fraction of nodes removed. In this work, we propose a strategy in\nwhich there is a pre-existing, dormant spare capacity already built into the\nnetwork for an identified vulnerable node, such that the traffic of the\ndisrupted node can be diverted to another pre-existing node/set of nodes in the\nnetwork. Using our algorithm, the increase in robustness of canonical\nscale-free networks was nearly 16-fold. We also analysed real-world networks\nusing our algorithm, where the mean increase in robustness was nearly\nfive-fold. To our knowledge, these numbers are significantly higher than those\nhitherto reported in literature. The normalized cost of this spare capacity and\nits effect on the operational parameters of the network have also been\ndiscussed. Instances of spare capacity in biological networks, termed as\ndistributed robustness, are also presented.\n", " This paper proposes a novel split learning framework with multiple\nend-systems in order to realize privacypreserving deep neural network\ncomputation. In conventional split learning frameworks, deep neural network\ncomputation is separated into multiple computing systems for hiding entire\nnetwork architectures. In our proposed framework, multiple computing\nend-systems are sharing one centralized server in split learning computation,\nwhere the multiple end-systems are with input and first hidden layers and the\ncentralized server is with the other hidden layers and output layer. This\nframework, which is called as spatio-temporal split learning, is spatially\nseparated for gathering data from multiple end-systems and also temporally\nseparated due to the nature of split learning. Our performance evaluation\nverifies that our proposed framework shows nearoptimal accuracy while\npreserving data privacy.\n", " Several adaptation techniques have been investigated to optimize fuzzy\ninference systems. Neural network learning algorithms have been used to\ndetermine the parameters of fuzzy inference system. Such models are often\ncalled as integrated neuro-fuzzy models. In an integrated neuro-fuzzy model\nthere is no guarantee that the neural network learning algorithm converges and\nthe tuning of fuzzy inference system will be successful. Success of\nevolutionary search procedures for optimization of fuzzy inference system is\nwell proven and established in many application areas. In this paper, we will\nexplore how the optimization of fuzzy inference systems could be further\nimproved using a meta-heuristic approach combining neural network learning and\nevolutionary computation. The proposed technique could be considered as a\nmethodology to integrate neural networks, fuzzy inference systems and\nevolutionary search procedures. We present the theoretical frameworks and some\nexperimental results to demonstrate the efficiency of the proposed technique.\n", " In this work, we explore the approximation capability of deep Rectified\nQuadratic Unit neural networks for H\\\"older-regular functions, with respect to\nthe uniform norm. We find that theoretical approximation heavily depends on the\nselected activation function in the neural network.\n", " Neural network training is usually accomplished by solving a non-convex\noptimization problem using stochastic gradient descent. Although one optimizes\nover the networks parameters, the main loss function generally only depends on\nthe realization of the neural network, i.e. the function it computes. Studying\nthe optimization problem over the space of realizations opens up new ways to\nunderstand neural network training. In particular, usual loss functions like\nmean squared error and categorical cross entropy are convex on spaces of neural\nnetwork realizations, which themselves are non-convex. Approximation\ncapabilities of neural networks can be used to deal with the latter\nnon-convexity, which allows us to establish that for sufficiently large\nnetworks local minima of a regularized optimization problem on the realization\nspace are almost optimal. Note, however, that each realization has many\ndifferent, possibly degenerate, parametrizations. In particular, a local\nminimum in the parametrization space needs not correspond to a local minimum in\nthe realization space. To establish such a connection, inverse stability of the\nrealization map is required, meaning that proximity of realizations must imply\nproximity of corresponding parametrizations. We present pathologies which\nprevent inverse stability in general, and, for shallow networks, proceed to\nestablish a restricted space of parametrizations on which we have inverse\nstability w.r.t. to a Sobolev norm. Furthermore, we show that by optimizing\nover such restricted sets, it is still possible to learn any function which can\nbe learned by optimization over unrestricted sets.\n", " Despite its better bio-plausibility, goal-driven spiking neural network (SNN)\nhas not achieved applicable performance for classifying biological spike\ntrains, and showed little bio-functional similarities compared to traditional\nartificial neural networks. In this study, we proposed the motorSRNN, a\nrecurrent SNN topologically inspired by the neural motor circuit of primates.\nBy employing the motorSRNN in decoding spike trains from the primary motor\ncortex of monkeys, we achieved a good balance between classification accuracy\nand energy consumption. The motorSRNN communicated with the input by capturing\nand cultivating more cosine-tuning, an essential property of neurons in the\nmotor cortex, and maintained its stability during training. Such\ntraining-induced cultivation and persistency of cosine-tuning was also observed\nin our monkeys. Moreover, the motorSRNN produced additional bio-functional\nsimilarities at the single-neuron, population, and circuit levels,\ndemonstrating biological authenticity. Thereby, ablation studies on motorSRNN\nhave suggested long-term stable feedback synapses contribute to the\ntraining-induced cultivation in the motor cortex. Besides these novel findings\nand predictions, we offer a new framework for building authentic models of\nneural computation.\n", " Implicit Neural Representations (INR) or neural fields have emerged as a\npopular framework to encode multimedia signals such as images and radiance\nfields while retaining high-quality. Recently, learnable feature grids proposed\nby Instant-NGP have allowed significant speed-up in the training as well as the\nsampling of INRs by replacing a large neural network with a multi-resolution\nlook-up table of feature vectors and a much smaller neural network. However,\nthese feature grids come at the expense of large memory consumption which can\nbe a bottleneck for storage and streaming applications. In this work, we\npropose SHACIRA, a simple yet effective task-agnostic framework for compressing\nsuch feature grids with no additional post-hoc pruning/quantization stages. We\nreparameterize feature grids with quantized latent weights and apply entropy\nregularization in the latent space to achieve high levels of compression across\nvarious domains. Quantitative and qualitative results on diverse datasets\nconsisting of images, videos, and radiance fields, show that our approach\noutperforms existing INR approaches without the need for any large datasets or\ndomain-specific heuristics. Our project page is available at\nhttp://shacira.github.io .\n", " Gene essentiality refers to the degree to which a gene is necessary for the\nsurvival and reproductive efficacy of a living organism. Although the\nessentiality of non-coding genes has been documented, there are still aspects\nof non-coding genes' essentiality that are unknown to us. For example, We do\nnot know the contribution of sequence features and network spatial features to\nessentiality. As a consequence, in this work, we propose DeepHEN that could\nanswer the above question. By buidling a new lncRNA-proteion-protein network\nand utilizing both representation learning and graph neural network, we\nsuccessfully build our DeepHEN models that could predict the essentiality of\nlncRNA genes. Compared to other methods for predicting the essentiality of\nlncRNA genes, our DeepHEN model not only tells whether sequence features or\nnetwork spatial features have a greater influence on essentiality but also\naddresses the overfitting issue of those methods caused by the low number of\nessential lncRNA genes, as evidenced by the results of enrichment analysis.\n", " Graph Neural Networks (GNNs) have emerged as a powerful and flexible\nframework for representation learning on irregular data. As they generalize the\noperations of classical CNNs on grids to arbitrary topologies, GNNs also bring\nmuch of the implementation challenges of their Euclidean counterparts. Model\nsize, memory footprint, and energy consumption are common concerns for many\nreal-world applications. Network binarization allocates a single bit to\nparameters and activations, thus dramatically reducing the memory requirements\n(up to 32x compared to single-precision floating-point numbers) and maximizing\nthe benefits of fast SIMD instructions on modern hardware for measurable\nspeedups. However, in spite of the large body of work on binarization for\nclassical CNNs, this area remains largely unexplored in geometric deep\nlearning. In this paper, we present and evaluate different strategies for the\nbinarization of graph neural networks. We show that through careful design of\nthe models, and control of the training process, binary graph neural networks\ncan be trained at only a moderate cost in accuracy on challenging benchmarks.\nIn particular, we present the first dynamic graph neural network in Hamming\nspace, able to leverage efficient k-NN search on binary vectors to speed-up the\nconstruction of the dynamic graph. We further verify that the binary models\noffer significant savings on embedded devices. Our code is publicly available\non Github.\n", " Bayesian neural learning feature a rigorous approach to estimation and\nuncertainty quantification via the posterior distribution of weights that\nrepresent knowledge of the neural network. This not only provides point\nestimates of optimal set of weights but also the ability to quantify\nuncertainty in decision making using the posterior distribution. Markov chain\nMonte Carlo (MCMC) techniques are typically used to obtain sample-based\nestimates of the posterior distribution. However, these techniques face\nchallenges in convergence and scalability, particularly in settings with large\ndatasets and network architectures. This paper address these challenges in two\nways. First, parallel tempering is used used to explore multiple modes of the\nposterior distribution and implemented in multi-core computing architecture.\nSecond, we make within-chain sampling schemes more efficient by using Langevin\ngradient information in forming Metropolis-Hastings proposal distributions. We\ndemonstrate the techniques using time series prediction and pattern\nclassification applications. The results show that the method not only improves\nthe computational time, but provides better prediction or decision making\ncapabilities when compared to related methods.\n", " Chemical reactions occur in energy, environmental, biological, and many other\nnatural systems, and the inference of the reaction networks is essential to\nunderstand and design the chemical processes in engineering and life sciences.\nYet, revealing the reaction pathways for complex systems and processes is still\nchallenging due to the lack of knowledge of the involved species and reactions.\nHere, we present a neural network approach that autonomously discovers reaction\npathways from the time-resolved species concentration data. The proposed\nChemical Reaction Neural Network (CRNN), by design, satisfies the fundamental\nphysics laws, including the Law of Mass Action and the Arrhenius Law.\nConsequently, the CRNN is physically interpretable such that the reaction\npathways can be interpreted, and the kinetic parameters can be quantified\nsimultaneously from the weights of the neural network. The inference of the\nchemical pathways is accomplished by training the CRNN with species\nconcentration data via stochastic gradient descent. We demonstrate the\nsuccessful implementations and the robustness of the approach in elucidating\nthe chemical reaction pathways of several chemical engineering and biochemical\nsystems. The autonomous inference by the CRNN approach precludes the need for\nexpert knowledge in proposing candidate networks and addresses the curse of\ndimensionality in complex systems. The physical interpretability also makes the\nCRNN capable of not only fitting the data for a given system but also\ndeveloping knowledge of unknown pathways that could be generalized to similar\nchemical systems.\n", " We demonstrate that the classification of boosted, hadronically-decaying weak\ngauge bosons can be significantly improved over traditional cut-based and\nBDT-based methods using deep learning and the jet charge variable. We construct\nbinary taggers for $W^+$ vs. $W^-$ and $Z$ vs. $W$ discrimination, as well as\nan overall ternary classifier for $W^+$/$W^-$/$Z$ discrimination. Besides a\nsimple convolutional neural network (CNN), we also explore a composite of two\nCNNs, with different numbers of layers in the jet $p_{T}$ and jet charge\nchannels. We find that this novel structure boosts the performance particularly\nwhen considering the $Z$ boson as signal. The methods presented here can\nenhance the physics potential in SM measurements and searches for new physics\nthat are sensitive to the electric charge of weak gauge bosons.\n", " The ribosome flow model with input and output (RFMIO) is a deterministic\ndynamical system that has been used to study the flow of ribosomes during mRNA\ntranslation. The RFMIO and its variants encapsulate important properties that\nare relevant to modeling ribosome flow such as the possible evolution of\n\"traffic jams\" and non-homogeneous elongation rates along the mRNA molecule,\nand can also be used for studying additional intracellular processes such as\ntranscription, transport, and more. Here we consider networks of interconnected\nRFMIOs as a fundamental tool for modeling, analyzing and re-engineering the\ncomplex mechanisms of protein production. In these networks, the output of each\nRFMIO may be divided, using connection weights, between several inputs of other\nRFMIOs. We show that under quite general feedback connections the network has\ntwo important properties: (1) it admits a unique steady-state and every\ntrajectory converges to this steady-state, and (2) the problem of how to\ndetermine the connection weights so that the network steady-state output is\nmaximized is a convex optimization problem. These mathematical properties make\nthese networks highly suitable as models of various phenomena: property (1)\nmeans that the behavior is predictable and ordered, and property (2) means that\ndetermining the optimal weights is numerically tractable even for large-scale\nnetworks. For the specific case of a feed-forward network of RFMIOs we prove an\nadditional useful property, namely, that there exists a spectral representation\nfor the network steady-state, and thus it can be determined without any\nnumerical simulations of the dynamics. We describe the implications of these\nresult to several fundamental biological phenomena and biotechnological\nobjectives.\n", " In this paper we study the theory of the so-called Kantorovich max-product\nneural network operators in the setting of Orlicz spaces $L^{\\varphi}$. The\nresults here proved, extend those given by Costarelli and Vinti in Result\nMath., 2016, to a more general context. The main advantage in studying neural\nnetwork type operators in Orlicz spaces relies in the possibility to\napproximate not necessarily continuous functions (data) belonging to different\nfunction spaces by a unique general approach. Further, in order to derive\nquantitative estimates in this context, we introduce a suitable K-functional in\n$L^{\\varphi}$ and use it to provide an upper bound for the approximation error\nof the above operators. Finally, examples of sigmoidal activation functions\nhave been considered and studied in details.\n", " Modern Neural Network (NN) architectures heavily rely on vast numbers of\nmultiply-accumulate arithmetic operations, constituting the predominant\ncomputational cost. Therefore, this paper proposes a high-throughput, scalable\nand energy efficient non-element-wise matrix multiplication unit on FPGAs as a\nbasic component of the NNs. We firstly streamline inter-layer and intra-layer\nredundancies of MADDNESS algorithm, a LUT-based approximate matrix\nmultiplication, to design a fast, efficient scalable approximate matrix\nmultiplication module termed \"Approximate Multiplication Unit (AMU)\". The AMU\noptimizes LUT-based matrix multiplications further through dedicated memory\nmanagement and access design, decoupling computational overhead from input\nresolution and boosting FPGA-based NN accelerator efficiency significantly. The\nexperimental results show that using our AMU achieves up to 9x higher\nthroughput and 112x higher energy efficiency over the state-of-the-art\nsolutions for the FPGA-based Quantised Neural Network (QNN) accelerators.\n", " A real-world text corpus sometimes comprises not only text documents but also\nsemantic links between them (e.g., academic papers in a bibliographic network\nare linked by citations and co-authorships). Text documents and semantic\nconnections form a text-rich network, which empowers a wide range of downstream\ntasks such as classification and retrieval. However, pretraining methods for\nsuch structures are still lacking, making it difficult to build one generic\nmodel that can be adapted to various tasks on text-rich networks. Current\npretraining objectives, such as masked language modeling, purely model texts\nand do not take inter-document structure information into consideration. To\nthis end, we propose our PretrAining on TexT-Rich NetwOrk framework Patton.\nPatton includes two pretraining strategies: network-contextualized masked\nlanguage modeling and masked node prediction, to capture the inherent\ndependency between textual attributes and network structure. We conduct\nexperiments on four downstream tasks in five datasets from both academic and\ne-commerce domains, where Patton outperforms baselines significantly and\nconsistently.\n", " In recent years, the scientific community has focused on the development of\nCAD tools that could improve bone fractures' classification, mostly based on\nConvolutional Neural Network (CNN). However, the discerning accuracy of\nfractures' subtypes was far from optimal. This paper proposes a modified\nversion of a very recent and powerful deep learning technique, the Vision\nTransformer (ViT), outperforming CNNs based approaches and consequently\nincreasing specialists' diagnosis accuracy. 4207 manually annotated images were\nused and distributed, by following the AO/OTA classification, in different\nfracture types, the largest labeled dataset of proximal femur fractures used in\nliterature. The ViT architecture was used and compared with a classic CNN and a\nmultistage architecture composed of successive CNNs in cascade. To demonstrate\nthe reliability of this approach, 1) the attention maps were used to visualize\nthe most relevant areas of the images, 2) the performance of a generic CNN and\nViT was compared through unsupervised learning techniques, and 3) 11\nspecialists were asked to evaluate and classify 150 proximal femur fractures'\nimages with and without the help of the ViT, then results were compared for\npotential improvement. The ViT was able to correctly predict 83% of the test\nimages. Precision, recall and F1-score were 0.77 (CI 0.64-0.90), 0.76 (CI\n0.62-0.91) and 0.77 (CI 0.64-0.89), respectively. The average specialists'\ndiagnostic improvement was 29% when supported by ViT's predictions,\noutperforming the algorithm alone. This paper showed the potential of Vision\nTransformers in bone fracture classification. For the first time, good results\nwere obtained in sub-fractures classification, with the largest and richest\ndataset ever. Accordingly, the assisted diagnosis yielded the best results,\nproving once again the effectiveness of a coordinated work between neural\nnetworks and specialists.\n", " We propose CoNSAL (Combining Neural networks and Symbolic regression for\nAnalytical Lyapunov function) to construct analytical Lyapunov functions for\nnonlinear dynamic systems. This framework contains a neural Lyapunov function\nand a symbolic regression component, where symbolic regression is applied to\ndistill the neural network to precise analytical forms. Our approach utilizes\nsymbolic regression not only as a tool for translation but also as a means to\nuncover counterexamples. This procedure terminates when no counterexamples are\nfound in the analytical formulation. Compared with previous results, our\nalgorithm directly produces an analytical form of the Lyapunov function with\nimproved interpretability in both the learning process and the final results.\nWe apply our algorithm to 2-D inverted pendulum, path following, Van Der Pol\nOscillator, 3-D trig dynamics, 4-D rotating wheel pendulum, 6-D 3-bus power\nsystem, and demonstrate that our algorithm successfully finds their valid\nLyapunov functions.\n", " We seek to augment a geometric network in the Euclidean plane with shortcuts\nto minimize its continuous diameter, i.e., the largest network distance between\nany two points on the augmented network. Unlike in the discrete setting where a\nshortcut connects two vertices and the diameter is measured between vertices,\nwe take all points along the edges of the network into account when placing a\nshortcut and when measuring distances in the augmented network.\n We study this network augmentation problem for paths and cycles. For paths,\nwe determine an optimal shortcut in linear time. For cycles, we show that a\nsingle shortcut never decreases the continuous diameter and that two shortcuts\nalways suffice to reduce the continuous diameter. Furthermore, we characterize\noptimal pairs of shortcuts for convex and non-convex cycles. Finally, we\ndevelop a linear time algorithm that produces an optimal pair of shortcuts for\nconvex cycles. Apart from the algorithms, our results extend to rectifiable\ncurves.\n Our work reveals some of the underlying challenges that must be overcome when\naddressing the discrete version of this network augmentation problem, where we\nminimize the discrete diameter of a network with shortcuts that connect only\nvertices.\n", " The case difference heuristic (CDH) approach is a knowledge-light method for\nlearning case adaptation knowledge from the case base of a case-based reasoning\nsystem. Given a pair of cases, the CDH approach attributes the difference in\ntheir solutions to the difference in the problems they solve, and generates\nadaptation rules to adjust solutions accordingly when a retrieved case and new\nquery have similar problem differences. As an alternative to learning\nadaptation rules, several researchers have applied neural networks to learn to\npredict solution differences from problem differences. Previous work on such\napproaches has assumed that the feature set describing problems is predefined.\nThis paper investigates a two-phase process combining deep learning for feature\nextraction and neural network based adaptation learning from extracted\nfeatures. Its performance is demonstrated in a regression task on an image\ndata: predicting age given the image of a face. Results show that the combined\nprocess can successfully learn adaptation knowledge applicable to nonsymbolic\ndifferences in cases. The CBR system achieves slightly lower performance\noverall than a baseline deep network regressor, but better performance than the\nbaseline on novel queries.\n", " With the growing amount of available temporal real-world network data, an\nimportant question is how to efficiently study these data. One can simply model\na temporal network as either a single aggregate static network, or as a series\nof time-specific snapshots, each of which is an aggregate static network over\nthe corresponding time window. The advantage of modeling the temporal data in\nthese two ways is that one can use existing well established methods for static\nnetwork analysis to study the resulting aggregate network(s). Here, we develop\na novel approach for studying temporal network data more explicitly. We base\nour methodology on the well established notion of graphlets (subgraphs), which\nhave been successfully used in numerous contexts in static network research.\nHere, we take the notion of static graphlets to the next level and develop new\ntheory needed to allow for graphlet-based analysis of temporal networks. Our\nnew notion of dynamic graphlets is quite different than existing approaches for\ndynamic network analysis that are based on temporal motifs (statistically\nsignificant subgraphs). Namely, these approaches suffer from many limitations.\nFor example, they can only deal with subgraph structures of limited complexity.\nAlso, their major drawback is that their results heavily depend on the choice\nof a null network model that is required to evaluate the significance of a\nsubgraph. However, choosing an appropriate null network model is a non-trivial\ntask. Our dynamic graphlet approach overcomes the limitations of the existing\ntemporal motif-based approaches. At the same time, when we thoroughly evaluate\nthe ability of our new approach to characterize the structure and function of\nan entire temporal network or of individual nodes, we find that the dynamic\ngraphlet approach outperforms the static graphlet approach, which indicates\nthat accounting for temporal information helps.\n", " In this work, we discuss use of machine learning techniques for rapid\nprediction of detonation properties including explosive energy, detonation\nvelocity, and detonation pressure. Further, analysis is applied to individual\nmolecules in order to explore the contribution of bonding motifs to these\nproperties. Feature descriptors evaluated include Morgan fingerprints, E-state\nvectors, a custom \"sum over bonds\" descriptor, and coulomb matrices. Algorithms\ndiscussed include kernel ridge regression, least absolute shrinkage and\nselection operator (\"LASSO\") regression, Gaussian process regression, and the\nmulti-layer perceptron (a neural network). Effects of regularization, kernel\nselection, network parameters, and dimensionality reduction are discussed. We\ndetermine that even when using a small training set, non-linear regression\nmethods may create models within a useful error tolerance for screening of\nmaterials.\n", " Understanding deep neural networks is a major research objective with notable\nexperimental and theoretical attention in recent years. The practical success\nof excessively large networks underscores the need for better theoretical\nanalyses and justifications. In this paper we focus on layer-wise functional\nstructure and behavior in overparameterized deep models. To do so, we study\nempirically the layers' robustness to post-training re-initialization and\nre-randomization of the parameters. We provide experimental results which give\nevidence for the heterogeneity of layers. Morally, layers of large deep neural\nnetworks can be categorized as either \"robust\" or \"critical\". Resetting the\nrobust layers to their initial values does not result in adverse decline in\nperformance. In many cases, robust layers hardly change throughout training. In\ncontrast, re-initializing critical layers vastly degrades the performance of\nthe network with test error essentially dropping to random guesses. Our study\nprovides further evidence that mere parameter counting or norm calculations are\ntoo coarse in studying generalization of deep models, and \"flatness\" and\nrobustness analysis of trained models need to be examined while taking into\naccount the respective network architectures.\n", " Time series analysis is a field of data science which is interested in\nanalyzing sequences of numerical values ordered in time. Time series are\nparticularly interesting because they allow us to visualize and understand the\nevolution of a process over time. Their analysis can reveal trends,\nrelationships and similarities across the data. There exists numerous fields\ncontaining data in the form of time series: health care (electrocardiogram,\nblood sugar, etc.), activity recognition, remote sensing, finance (stock market\nprice), industry (sensors), etc. Time series classification consists of\nconstructing algorithms dedicated to automatically label time series data. The\nsequential aspect of time series data requires the development of algorithms\nthat are able to harness this temporal property, thus making the existing\noff-the-shelf machine learning models for traditional tabular data suboptimal\nfor solving the underlying task. In this context, deep learning has emerged in\nrecent years as one of the most effective methods for tackling the supervised\nclassification task, particularly in the field of computer vision. The main\nobjective of this thesis was to study and develop deep neural networks\nspecifically constructed for the classification of time series data. We thus\ncarried out the first large scale experimental study allowing us to compare the\nexisting deep methods and to position them compared other non-deep learning\nbased state-of-the-art methods. Subsequently, we made numerous contributions in\nthis area, notably in the context of transfer learning, data augmentation,\nensembling and adversarial attacks. Finally, we have also proposed a novel\narchitecture, based on the famous Inception network (Google), which ranks among\nthe most efficient to date.\n", " Traffic signal control (TSC) is a challenging problem within intelligent\ntransportation systems and has been tackled using multi-agent reinforcement\nlearning (MARL). While centralized approaches are often infeasible for\nlarge-scale TSC problems, decentralized approaches provide scalability but\nintroduce new challenges, such as partial observability. Communication plays a\ncritical role in decentralized MARL, as agents must learn to exchange\ninformation using messages to better understand the system and achieve\neffective coordination. Deep MARL has been used to enable inter-agent\ncommunication by learning communication protocols in a differentiable manner.\nHowever, many deep MARL communication frameworks proposed for TSC allow agents\nto communicate with all other agents at all times, which can add to the\nexisting noise in the system and degrade overall performance. In this study, we\npropose a communication-based MARL framework for large-scale TSC. Our framework\nallows each agent to learn a communication policy that dictates \"which\" part of\nthe message is sent \"to whom\". In essence, our framework enables agents to\nselectively choose the recipients of their messages and exchange variable\nlength messages with them. This results in a decentralized and flexible\ncommunication mechanism in which agents can effectively use the communication\nchannel only when necessary. We designed two networks, a synthetic $4 \\times 4$\ngrid network and a real-world network based on the Pasubio neighborhood in\nBologna. Our framework achieved the lowest network congestion compared to\nrelated methods, with agents utilizing $\\sim 47-65 \\%$ of the communication\nchannel. Ablation studies further demonstrated the effectiveness of the\ncommunication policies learned within our framework.\n", " Electromagnetic (EM) body models designed to predict Radio-Frequency (RF)\npropagation are time-consuming methods which prevent their adoption in strict\nreal-time computational imaging problems, such as human body localization and\nsensing. Physics-informed Generative Neural Network (GNN) models have been\nrecently proposed to reproduce EM effects, namely to simulate or reconstruct\nmissing data or samples by incorporating relevant EM principles and\nconstraints. The paper discusses a Variational Auto-Encoder (VAE) model which\nis trained to reproduce the effects of human motions on the EM field and\nincorporate EM body diffraction principles. Proposed physics-informed\ngenerative neural network models are verified against both classical\ndiffraction-based EM tools and full-wave EM body simulations.\n", " Convolutional Neural Networks (CNN) are the state-of-the-art in the field of\nvisual computing. However, a major problem with CNNs is the large number of\nfloating point operations (FLOPs) required to perform convolutions for large\ninputs. When considering the application of CNNs to video data, convolutional\nfilters become even more complex due to the extra temporal dimension. This\nleads to problems when respective applications are to be deployed on mobile\ndevices, such as smart phones, tablets, micro-controllers or similar,\nindicating less computational power.\n Kim et al. (2016) proposed using a Tucker-decomposition to compress the\nconvolutional kernel of a pre-trained network for images in order to reduce the\ncomplexity of the network, i.e. the number of FLOPs. In this paper, we\ngeneralize the aforementioned method for application to videos (and other 3D\nsignals) and evaluate the proposed method on a modified version of the THETIS\ndata set, which contains videos of individuals performing tennis shots. We show\nthat the compressed network reaches comparable accuracy, while indicating a\nmemory compression by a factor of 51. However, the actual computational\nspeed-up (factor 1.4) does not meet our theoretically derived expectation\n(factor 6).\n", " In India financial markets have existed for many years. A functionally\naccented, diverse, efficient and flexible financial system is vital to the\nnational objective of creating a market driven, productive and competitive\neconomy. Today markets of varying maturity exist in equity, debt, commodities\nand foreign exchange. In this work we attempt to generate prediction rules\nscheme for stock price movement at Bombay Stock Exchange using an important\nSoft Computing paradigm viz., Rough Fuzzy Multi Layer Perception. The use of\nComputational Intelligence Systems such as Neural Networks, Fuzzy Sets, Genetic\nAlgorithms, etc. for Stock Market Predictions has been widely established. The\nprocess is to extract knowledge in the form of rules from daily stock\nmovements. These rules can then be used to guide investors. To increase the\nefficiency of the prediction process, Rough Sets is used to discretize the\ndata. The methodology uses a Genetic Algorithm to obtain a structured network\nsuitable for both classification and rule extraction. The modular concept,\nbased on divide and conquer strategy, provides accelerated training and a\ncompact network suitable for generating a minimum number of rules with high\ncertainty values. The concept of variable mutation operator is introduced for\npreserving the localized structure of the constituting Knowledge Based\nsub-networks, while they are integrated and evolved. Rough Set Dependency Rules\nare generated directly from the real valued attribute table containing Fuzzy\nmembership values. The paradigm is thus used to develop a rule extraction\nalgorithm. The extracted rules are compared with some of the related rule\nextraction techniques on the basis of some quantitative performance indices.\nThe proposed methodology extracts rules which are less in number, are accurate,\nhave high certainty factor and have low confusion with less computation time.\n", " Planar graph drawings tend to be aesthetically pleasing. In this poster we\nexplore a Neural Network's capability of learning various planar graph classes.\nAdditionally, we also investigate the effectiveness of the model in\ngeneralizing beyond planarity. We find that the model can outperform\nconventional techniques for certain graph classes. The model, however, appears\nto be more susceptible to randomness in the data, and seems to be less robust\nthan expected.\n", " A dynamical system obtains a wide variety of kinetic realizations, which is\nadvantageous for the analysis of biochemical systems. A reaction network,\nderived from a dynamical system, may or may not possess some properties needed\nfor a thorough analysis. We improve and extend the work of M. Johnston\n\\cite{JOHN2014} and Hong et al. \\cite{HONG2023} on network translations to\nnetwork transformations, where the network is modified while preserving the\ndynamical system. These transformations can shrink, extend, or retain the\nstoichiometric subspace. Here, we show that positive dependent network can be\ntranslated to a weakly reversible network. Using the kinetic realizations of\n(1) calcium signaling in the olfactory system and (2) metabolic insulin\nsignaling, we demonstrate the benefits of transformed systems with positive\ndeficiency for analyzing biochemical systems. Furthermore, we present an\nalgorithm for a network transformation of a weakly reversible non-complex\nfactorizable kinetic (NFK) system to a weakly reversible complex factorizable\nkinetic (CFK) system, thereby enhancing the Subspace Coincidence Theorem for\nNFK systems of Nazareno et al. \\cite{NAZA2019}. Finally, using the transformed\nkinetic realization of monolignol biosynthesis in \\textit{Populus xylem}, we\nstudy the structural and kinetic properties of transformed systems,including\nthe invariance of concordance and variation of injectivity and\nmono-/multi-stationarity under network transformation.\n", " Neural networks functions are supposed to be able to encode the desired\nsolution of an inverse problem very efficiently. In this paper, we consider the\nproblem of solving linear inverse problems with neural network coders. First we\nestablish some correspondences of this formulation with existing concepts in\nregularization theory, in particular with state space regularization, operator\ndecomposition and iterative regularization methods. A Gauss-Newton's method is\nsuitable for solving encoded linear inverse problems, which is supported by a\nlocal convergence result. The convergence studies, however, are not complete,\nand are based on a conjecture on linear independence of activation functions\nand its derivatives.\n", " Reward decomposition is a critical problem in centralized training with\ndecentralized execution~(CTDE) paradigm for multi-agent reinforcement learning.\nTo take full advantage of global information, which exploits the states from\nall agents and the related environment for decomposing Q values into individual\ncredits, we propose a general meta-learning-based Mixing Network with Meta\nPolicy Gradient~(MNMPG) framework to distill the global hierarchy for delicate\nreward decomposition. The excitation signal for learning global hierarchy is\ndeduced from the episode reward difference between before and after \"exercise\nupdates\" through the utility network. Our method is generally applicable to the\nCTDE method using a monotonic mixing network. Experiments on the StarCraft II\nmicromanagement benchmark demonstrate that our method just with a simple\nutility network is able to outperform the current state-of-the-art MARL\nalgorithms on 4 of 5 super hard scenarios. Better performance can be further\nachieved when combined with a role-based utility network.\n", " This study presents a generalization for a method examining the correlation\nfunction of an arbitrary system with interactions in an Ising model to obtain a\nvalue of correlation between two arbitrary points on a network. The\nestablishment of a network clarifies the type of calculations necessary for the\ncorrelation values between secondary and tertiary nodes. Moreover, it is\npossible to calculate the correlation values of the target that are interlinked\nin a complex manner by proposing a network analysis method to express the same\nas a network with mutual linkages between the target of each field.\n", " Graph Visualization, also known as Graph Drawing, aims to find geometric\nembeddings of graphs that optimize certain criteria. Stress is a widely used\nmetric; stress is minimized when every pair of nodes is positioned at their\nshortest path distance. However, stress optimization presents computational\nchallenges due to its inherent complexity and is usually solved using\nheuristics in practice. We introduce a scalable Graph Neural Network (GNN)\nbased Graph Drawing framework with sub-quadratic runtime that can learn to\noptimize stress. Inspired by classical stress optimization techniques and\nforce-directed layout algorithms, we create a coarsening hierarchy for the\ninput graph. Beginning at the coarsest level, we iteratively refine and\nun-coarsen the layout, until we generate an embedding for the original graph.\nTo enhance information propagation within the network, we propose a novel\npositional rewiring technique based on intermediate node positions. Our\nempirical evaluation demonstrates that the framework achieves state-of-the-art\nperformance while remaining scalable.\n", " Data-driven dynamic models of cell biology can be used to predict cell\nresponse to unseen perturbations. Recent work (CellBox) had demonstrated the\nderivation of interpretable models with explicit interaction terms, in which\nthe parameters were optimized using machine learning techniques. While the\nprevious work was tested only in a single biological setting, this work aims to\nextend the range of applicability of this model inference approach to a\ndiversity of biological systems. Here we adapted CellBox in Julia differential\nprogramming and augmented the method with adjoint algorithms, which has\nrecently been used in the context of neural ODEs. We trained the models using\nsimulated data from both abstract and biology-inspired networks, which afford\nthe ability to evaluate the recovery of the ground truth network structure. The\nresulting accuracy of prediction by these models is high both in terms of low\nerror against data and excellent agreement with the network structure used for\nthe simulated training data. While there is no analogous ground truth for real\nlife biological systems, this work demonstrates the ability to construct and\nparameterize a considerable diversity of network models with high predictive\nability. The expectation is that this kind of procedure can be used on real\nperturbation-response data to derive models applicable to diverse biological\nsystems.\n", " Biological neural networks are characterized by their high degree of\nplasticity, a core property that enables the remarkable adaptability of natural\norganisms. Importantly, this ability affects both the synaptic strength and the\ntopology of the nervous systems. Artificial neural networks, on the other hand,\nhave been mainly designed as static, fully connected structures that can be\nnotoriously brittle in the face of changing environments and novel inputs.\nBuilding on previous works on Neural Developmental Programs (NDPs), we propose\na class of self-organizing neural networks capable of synaptic and structural\nplasticity in an activity and reward-dependent manner which we call Lifelong\nNeural Developmental Program (LNDP). We present an instance of such a network\nbuilt on the graph transformer architecture and propose a mechanism for\npre-experience plasticity based on the spontaneous activity of sensory neurons.\nOur results demonstrate the ability of the model to learn from experiences in\ndifferent control tasks starting from randomly connected or empty networks. We\nfurther show that structural plasticity is advantageous in environments\nnecessitating fast adaptation or with non-stationary rewards.\n", " Real-time traffic prediction models play a pivotal role in smart mobility\nsystems and have been widely used in route guidance, emerging mobility\nservices, and advanced traffic management systems. With the availability of\nmassive traffic data, neural network-based deep learning methods, especially\nthe graph convolutional networks (GCN) have demonstrated outstanding\nperformance in mining spatio-temporal information and achieving high prediction\naccuracy. Recent studies reveal the vulnerability of GCN under adversarial\nattacks, while there is a lack of studies to understand the vulnerability\nissues of the GCN-based traffic prediction models. Given this, this paper\nproposes a new task -- diffusion attack, to study the robustness of GCN-based\ntraffic prediction models. The diffusion attack aims to select and attack a\nsmall set of nodes to degrade the performance of the entire prediction model.\nTo conduct the diffusion attack, we propose a novel attack algorithm, which\nconsists of two major components: 1) approximating the gradient of the\nblack-box prediction model with Simultaneous Perturbation Stochastic\nApproximation (SPSA); 2) adapting the knapsack greedy algorithm to select the\nattack nodes. The proposed algorithm is examined with three GCN-based traffic\nprediction models: St-Gcn, T-Gcn, and A3t-Gcn on two cities. The proposed\nalgorithm demonstrates high efficiency in the adversarial attack tasks under\nvarious scenarios, and it can still generate adversarial samples under the drop\nregularization such as DropOut, DropNode, and DropEdge. The research outcomes\ncould help to improve the robustness of the GCN-based traffic prediction models\nand better protect the smart mobility systems. Our code is available at\nhttps://github.com/LYZ98/Adversarial-Diffusion-Attacks-on-Graph-based-Traffic-Prediction-Models\n", " Given a n points in two dimensional space, a Manhattan Network G is a network\nthat connects all n points with either horizontal or vertical edges, with the\nproperty that for any two point in G should be connected by a Manhattan path\nand distance between this two points is equal to Manhattan Distance. The\nMinimum Manhattan Network problem is to find a Manhattan network with minimum\nnetwork length, i.e., summation of all line segment in network should be\nminimize. In this paper, we proposed a 2-approximation algorithm with time\ncomplexity O(|E|lgN) where |E| is the number of edges and N is the number of\nnodes. Using randomly generated datasets, we compare our result with the\noptimal one.\n", " This paper explores the problem of effectively compressing 3D geometry sets\ncontaining diverse categories. We make \\textit{the first} attempt to tackle\nthis fundamental and challenging problem and propose NeCGS, a neural\ncompression paradigm, which can compress hundreds of detailed and diverse 3D\nmesh models (~684 MB) by about 900 times (0.76 MB) with high accuracy and\npreservation of detailed geometric details. Specifically, we first represent\neach irregular mesh model/shape in a regular representation that implicitly\ndescribes the geometry structure of the model using a 4D regular volume, called\nTSDF-Def volume. Such a regular representation can not only capture local\nsurfaces more effectively but also facilitate the subsequent process. Then we\nconstruct a quantization-aware auto-decoder network architecture to regress\nthese 4D volumes, which can summarize the similarity of local geometric\nstructures within a model and across different models for redundancy\nlimination, resulting in more compact representations, including an embedded\nfeature of a smaller size associated with each model and a network parameter\nset shared by all models. We finally encode the resulting features and network\nparameters into bitstreams through entropy coding. After decompressing the\nfeatures and network parameters, we can reconstruct the TSDF-Def volumes, where\nthe 3D surfaces can be extracted through the deformable marching\ncubes.Extensive experiments and ablation studies demonstrate the significant\nadvantages of our NeCGS over state-of-the-art methods both quantitatively and\nqualitatively.\n", " Sparseness is a useful regularizer for learning in a wide range of\napplications, in particular in neural networks. This paper proposes a model\ntargeted at classification tasks, where sparse activity and sparse connectivity\nare used to enhance classification capabilities. The tool for achieving this is\na sparseness-enforcing projection operator which finds the closest vector with\na pre-defined sparseness for any given vector. In the theoretical part of this\npaper, a comprehensive theory for such a projection is developed. In\nconclusion, it is shown that the projection is differentiable almost everywhere\nand can thus be implemented as a smooth neuronal transfer function. The entire\nmodel can hence be tuned end-to-end using gradient-based methods. Experiments\non the MNIST database of handwritten digits show that classification\nperformance can be boosted by sparse activity or sparse connectivity. With a\ncombination of both, performance can be significantly better compared to\nclassical non-sparse approaches.\n", " The number of fixed mutations accumulated in an evolving population often\ndisplays a variance that is significantly larger than the mean (the\noverdispersed molecular clock). By examining a generic evolutionary process on\na neutral network of high-fitness genotypes, we establish a formalism for\ncomputing all cumulants of the full probability distribution of accumulated\nmutations in terms of graph properties of the neutral network, and use the\nformalism to prove overdispersion of the molecular clock. We further show that\nsignificant overdispersion arises naturally in evolution when the neutral\nnetwork is highly sparse, exhibits large global fluctuations in neutrality, and\nsmall local fluctuations in neutrality. The results are also relevant for\nelucidating the topological structure of a neutral network from empirical\nmeasurements of the substitution process.\n", " Exploiting recent developments in information theory, we propose, illustrate,\nand validate a principled information-theoretic algorithm for module discovery\nand resulting measure of network modularity. This measure is an order parameter\n(a dimensionless number between 0 and 1). Comparison is made to other\napproaches to module-discovery and to quantifying network modularity using\nMonte Carlo generated Erdos-like modular networks. Finally, the Network\nInformation Bottleneck (NIB) algorithm is applied to a number of real world\nnetworks, including the \"social\" network of coauthors at the APS March Meeting\n2004.\n", " Networks of neurons in the brain encode preferred patterns of neural activity\nvia their synaptic connections. Despite receiving considerable attention, the\nprecise relationship between network connectivity and encoded patterns is still\npoorly understood. Here we consider this problem for networks of\nthreshold-linear neurons whose computational function is to learn and store a\nset of binary patterns (e.g., a neural code) as \"permitted sets\" of the\nnetwork. We introduce a simple Encoding Rule that selectively turns \"on\"\nsynapses between neurons that co-appear in one or more patterns. The rule uses\nsynapses that are binary, in the sense of having only two states (\"on\" or\n\"off\"), but also heterogeneous, with weights drawn from an underlying synaptic\nstrength matrix S. Our main results precisely describe the stored patterns that\nresult from the Encoding Rule -- including unintended \"spurious\" states -- and\ngive an explicit characterization of the dependence on S. In particular, we\nfind that binary patterns are successfully stored in these networks when the\nexcitatory connections between neurons are geometrically balanced -- i.e., they\nsatisfy a set of geometric constraints. Furthermore, we find that certain types\nof neural codes are \"natural\" in the context of these networks, meaning that\nthe full code can be accurately learned from a highly undersampled set of\npatterns. Interestingly, many commonly observed neural codes in cortical and\nhippocampal areas are natural in this sense. As an application, we construct\nnetworks that encode hippocampal place field codes nearly exactly, following\npresentation of only a small fraction of patterns. To obtain our results, we\nprove new theorems using classical ideas from convex and distance geometry,\nsuch as Cayley-Menger determinants, revealing a novel connection between these\nareas of mathematics and coding properties of neural networks.\n", " Protein representation learning methods have shown great potential to yield\nuseful representation for many downstream tasks, especially on protein\nclassification. Moreover, a few recent studies have shown great promise in\naddressing insufficient labels of proteins with self-supervised learning\nmethods. However, existing protein language models are usually pretrained on\nprotein sequences without considering the important protein structural\ninformation. To this end, we propose a novel structure-aware protein\nself-supervised learning method to effectively capture structural information\nof proteins. In particular, a well-designed graph neural network (GNN) model is\npretrained to preserve the protein structural information with self-supervised\ntasks from a pairwise residue distance perspective and a dihedral angle\nperspective, respectively. Furthermore, we propose to leverage the available\nprotein language model pretrained on protein sequences to enhance the\nself-supervised learning. Specifically, we identify the relation between the\nsequential information in the protein language model and the structural\ninformation in the specially designed GNN model via a novel pseudo bi-level\noptimization scheme. Experiments on several supervised downstream tasks verify\nthe effectiveness of our proposed method.The code of the proposed method is\navailable in \\url{https://github.com/GGchen1997/STEPS_Bioinformatics}.\n", " In this study, we have developed a parallel version of the random time\nsimulation algorithm.\n Firstly, we gave a rigorous basis of the random time description of the\nstochastic process of chemical reaction network time evolution.\n And then we reviewed the random time simulation algorithm and gave the\nimplementations for the parallel version of next reaction random time\nalgorithm.\n The discussion of computational complexity suggested a factor of $M$ (which\nis the connection number of the network) folds time consuming reduction for\nrandom time simulation algorithm as compared to other exact stochastic\nsimulation algorithms, such as the Gillespie algorithm.\n For large-scale system, such like the protein-protein interaction network,\n$M$ is on order of $10^8$.\n We further demonstrate the power of random time simulation with a GPGPU\nparallel implementation which achieved roughly 100 folds acceleration as\ncompared with CPU implementations.\n Therefore the stochastic simulation method we developed here can be of great\napplication value for simulating time evolution process of large-scale network.\n", " The use of network based approaches to model and analyse large datasets is\ncurrently a growing research field. For instance in biology and medicine,\nnetworks are used to model interactions among biological molecules as well as\nrelations among patients. Similarly, data coming from social networks can be\ntrivially modelled by using graphs. More recently, the use of dual networks\ngained the attention of researchers. A dual network model uses a pair of graphs\nto model a scenario in which one of the two graphs is usually unweighted (a\nnetwork representing physical associations among nodes) while the other one is\nedge-weighted (a network representing conceptual associations among nodes). In\nthis paper we focus on the problem of finding the Densest Connected sub-graph\n(DCS) having the largest density in the conceptual network which is also\nconnected in the physical network. The problem is relevant but also\ncomputationally hard, therefore the need for introducing of novel algorithms\narises. We formalise the problem and then we map DCS into a graph alignment\nproblem. Then we propose a possible solution. A set of experiments is also\npresented to support our approach.\n" ]

[ " We report some computational results comparing parallel and sequential codes\nfor vertex/facet enumeration problems for convex polyhedra. The problems chosen\nspan the range from simple to highly degenerate polytopes. We tested one code\n(lrs) based on pivoting and four codes (cddr+, ppl, normaliz, PORTA) based on\nthe double description method. normaliz employs parallelization as do the codes\nplrs and mplrs which are based on lrs. We tested these codes using various\nhardware configurations with up to 1200 cores. Major speedups were obtained by\nparallelization, particularly by the code mplrs which uses MPI and can operate\non clusters of machines.\n", " In this article we present a class of codes with few weights arising from\nspecial type of linear sets. We explicitly show the weights of such codes,\ntheir weight enumerator and possible choices for their generator matrices. In\nparticular, our construction yields also to linear codes with three weights\nand, in some cases, to almost MDS codes. The interest for these codes relies on\ntheir applications to authentication codes and secret schemes, and their\nconnections with further objects such as association schemes and graphs.\n", " In this paper we analyse the construction of identification codes.\nIdentification codes are based on the question \"Is the message I have just\nreceived the one I am interested in?\", as opposed to Shannon's transmission,\nwhere the receiver is interested in not only one, but any message. The\nadvantage of identification is that it allows rates growing double\nexponentially in the blocklength at the cost of not being able to decode every\nmessage, which might be beneficial in certain applications. We focus on a\nspecial identification code construction based on two concatenated Reed-Solomon\ncodes and have a closer look at its implementation, analyzing the trade-offs of\nidentification with respect to transmission and the trade-offs introduced by\nthe computational cost of identification codes.\n", " Using the ideas of E.I. Gordon we present and farther advance an approach,\nbased on nonstandard analysis, to simultaneous approximations of locally\ncompact abelian groups and their duals by (hyper)finite abelian groups, as well\nas to approximations of various types of Fourier transforms on them by the\ndiscrete Fourier transform. Combining some methods of nonstandard analysis and\nadditive combinatorics we prove the three Gordon's Conjectures which were open\nsince 1991 and are crucial both in the formulations and proofs of the LCA\ngroups and Fourier transform approximation theorems.\n", " We consider non-perturbative effects at short distances in theories with\nconfinement. The analysis is straightforward within the Abelian models in which\nthe confinement arises on classical level. In all cases considered (compact\nU(1) in 3D and 4D, dual Abelian Higgs model) there are non-perturbative\ncontributions associated with short distances which are due to topological\ndefects. In QCD case, both classical and quantum effects determine the role of\nthe topological defects and the theoretical analysis has not been completed so\nfar. Generically, the topological defects would result in 1/Q^2 corrections\ngoing beyond the standard Operator Product Expansion. We review existing data\non the power corrections and find that the data favor existence of the novel\ncorrections, at least at the mass scale of (1-2) GeV. We indicate crucial\nexperiments which could further clarify the situation on the phenomenological\nside.\n", " Convolutional codes are constructed, designed and analysed using row and/or\nblock structures of unit algebraic schemes. Infinite series of such codes and\nof codes with specific properties are derived. Properties are shown\nalgebraically and algebraic decoding methods are derived. For a given rate and\ngiven error-correction capability at each component, convolutional codes with\nthese specifications and with efficient decoding algorithms are constructed.\nExplicit prototype examples are given but in general large lengths and large\nerror capability are achievable. Convolutional codes with efficient decoding\nalgorithms at or near the maximum free distances attainable for the parameters\nare constructible. Unit memory convolutional codes of maximum possible free\ndistance are designed with practical algebraic decoding algorithms.\n LDPC (low density parity check) convolutional codes with efficient decoding\nschemes are constructed and analysed by the methods. Self-dual and\ndual-containing convolutional codes may also be designed by the methods;\ndual-containing codes enables the construction of quantum codes.\n", " The Riesz transform of $u$ : $\\mathcal{S}(\\mathbb{R}^n) \\rightarrow\n\\mathcal{S'}(\\mathbb{R}^n)$ is defined as a convolution by a singular kernel,\nand can be conveniently expressed using the Fourier Transform and a simple\nmultiplier. We extend this analysis to higher order Riesz transforms, i.e. some\ntype of singular integrals that contain tensorial polyadic kernels and define\nan integral transform for functions $\\mathcal{S}(\\mathbb{R}^n) \\rightarrow\n\\mathcal{S'}(\\mathbb{R}^{ n \\times n \\times \\dots n})$. We show that the\ntransformed kernel is also a polyadic tensor, and propose a general method to\ncompute explicitely the Fourier mutliplier. Analytical results are given, as\nwell as a recursive algorithm, to compute the coefficients of the transformed\nkernel. We compare the result to direct numerical evaluation, and discuss the\ncase $n=2$, with application to image analysis.\n", " The aim of this work is to give degree formulas for the generalized Hamming\nweights of evaluation codes and to show lower bounds for these weights. In\nparticular, we give degree formulas for the generalized Hamming weights of\nReed--Muller-type codes, and we determine the minimum distance of toric codes\nover hypersimplices, and the 1st and 2nd generalized Hamming weights of\nsquarefree evaluation codes.\n", " A new approach for the approximation of the channel log-likelihood ratio\n(LLR) for wireless channels based on Taylor series is proposed. The\napproximation is applied to the uncorrelated flat Rayleigh fading channel with\nunknown channel state information at the receiver. It is shown that the\nproposed approximation greatly simplifies the calculation of channel LLRs, and\nyet provides results almost identical to those based on the exact calculation\nof channel LLRs. The results are obtained in the context of bit-interleaved\ncoded modulation (BICM) schemes with low-density parity-check (LDPC) codes, and\ninclude threshold calculations and error rate performance of finite-length\ncodes. Compared to the existing approximations, the proposed method is either\nsignificantly less complex, or considerably more accurate.\n", " We present sufficient conditions for the existence of matchings in abelian\ngroups and their linear counterparts. These conditions lead to extensions of\nexisting results in matching theory. Additionally, we classify subsets within\nabelian groups that cannot be matched. We introduce the concept of Chowla\nsubspaces and formulate and conjecture a linear analogue of a result originally\nattributed to Y. O. Hamidoune [20] concerning Chowla sets. If proven true, this\nresult would extend matchings in primitive subspaces. Throughout the paper, we\nemphasize the analogy between matchings in abelian groups and field extensions.\nWe also pose numerous open questions for future research. Our approach relies\non classical theorems in group theory, additive number theory and linear\nalgebra. As the title of the paper suggests, this work is the second sequel to\na previous paper [5] with a similar theme. This paper is self-contained and can\nbe read independently.\n", " Huffman Codes are optimal Instantaneous Fixed-to-Variable (FV) codes in which\nevery source symbol can only be encoded by one codeword. Relaxing these\nconstraints permits constructing better FV codes. More specifically, recent\nwork has shown that AIFV-$m$ codes can beat Huffman coding. AIFV-$m$ codes\nconstruct am $m$-tuple of different coding trees between which the code\nalternates and are only almost instantaneous (AI). This means that decoding a\nword might require a delay of a finite number of bits.\n Current algorithms for constructing optimal AIFV-$m$ codes are iterative\nprocesses that construct progressively \"better sets\" of code trees. The\nprocesses have been proven to finitely converge to the optimal code but with no\nknown bounds on the convergence rate.\n This paper derives a geometric interpretation of the space of binary AIFV-$2$\ncodes, permitting the development of the first polynomially time-bounded\niterative procedures for constructing optimal AIFV codes.\n We first show that a simple binary search procedure can replace the current\niterative process to construct optimal binary AIFV-$2$ codes. We then describe\nhow to frame the problem as a linear programming with an exponential number of\nconstraints but a polynomial-time separability oracle. This permits using the\nGr\\\"otschel, Lov\\'asz and Schrijver ellipsoid method to solve the problem in a\npolynomial number of steps. While more complicated, this second method has the\npotential to lead to a polynomial time algorithm to construct optimal AIFV-$m$\ncodes for general $m$.\n", " We extend the Main Theorem of Aschbacher and Smith on Quillen's Conjecture\nfrom $p>5$ to the remaining odd primes $p = 3,5$. In the process, we develop\nfurther combinatorial and homotopical methods for studying the poset of\nnontrivial elementary abelian $p$-subgroups of a finite group. The techniques\nlead to a number of further results on the Conjecture, often reducing\ndependence on the CFSG; in particular, we also provide some partial results\ntoward the case of $p=2$.\n", " Previous work on convexity of neural codes has produced codes that are\nopen-convex but not closed-convex -- or vice-versa. However, why a code is one\nbut not the other, and how to detect such discrepancies are open questions. We\ntackle these questions in two ways. First, we investigate the concept of\ndegeneracy introduced by Cruz et al., and extend their results to show that\nnondegeneracy precisely captures the situation when taking closures or\ninteriors of open or closed realizations, respectively, yields another\nrealization of the code. Second, we give the first general criteria for\nprecluding a code from being closed-convex (without ruling out open-convexity),\nunifying ad-hoc geometric arguments in prior works. One criterion is built on a\nphenomenon we call a rigid structure, while the other can be stated\nalgebraically, in terms of the neural ideal of the code. These results\ncomplement existing criteria having the opposite purpose: precluding\nopen-convexity but not closed-convexity. Finally, we show that a family of\ncodes shown by Jeffs to be not open-convex is in fact closed-convex and\nrealizable in dimension two.\n", " We construct bijections giving three \"codes\" for trees. These codes follow\nnaturally from the Matrix Tree Theorem of Tutte and have many advantages over\nthe one produced by Prufer in 1918. One algorithm gives explicitly a bijection\nthat is implicit in Orlin's manipulatorial proof of Cayley's formula (the\nformula was actually found first by Borchardt). Another is based on a proof of\nKnuth. The third is an implementation of Joyal's pseudo-bijective proof of the\nformula, and is equivalent to one previously found by Egecioglu and Remmel. In\neach case, we have at least two algorithms, one of which involves hands-on\nmanipulations of the tree while the other involves a combinatorial and linear\nalgebraic manipulation of a matrix.\n", " This paper considers the degree-diameter problem for undirected circulant\ngraphs. For degrees 10 and 11 newly discovered families of circulant graphs of\narbitrary diameter are presented which are largest known and are conjectured to\nbe extremal. They are also the largest-known Abelian Cayley graphs of these\ndegrees. For each such family the order of every graph in the family is defined\nby a quintic polynomial function of the diameter which is specific to the\nfamily. The elements of the generating set for each graph are similarly defined\nby a set of polynomials in the diameter. The existence of the graphs in the\ndegree 10 families has been proved for all diameters. These graphs are\nconsistent with a conjecture on the order of extremal Abelian Cayley and\ncirculant graphs of any degree and diameter.\n This is the extended version of the paper, including the proof steps for\ndegree 10 graphs covering all diameter classes and an appendix listing\nadditional tables of generating sets.\n", " We present a unified theory for the almost periodicity of functions with\nvalues in an arbitrary Banach space, measures and distributions via almost\nperiodic elements for the action of a locally compact abelian group on a\nuniform topological space. We discuss the relation between Bohr and Bochner\ntype almost periodicity, and similar conditions, and how the equivalence among\nsuch conditions relates to properties of the group action and the uniformity.\nWe complete the paper by demonstrating how various examples considered earlier\nall fit in our framework.\n", " A new generalization of the Gray map is introduced. The new generalization\n$\\Phi: Z_{2^k}^n \\to Z_{2}^{2^{k-1}n}$ is connected with the known generalized\nGray map $\\phi$ in the following way: if we take two dual linear\n$Z_{2^k}$-codes and construct binary codes from them using the generalizations\n$\\phi$ and $\\Phi$ of the Gray map, then the weight enumerators of the binary\ncodes obtained will satisfy the MacWilliams identity. The classes of\n$Z_{2^k}$-linear Hadamard codes and co-$Z_{2^k}$-linear extended 1-perfect\ncodes are described, where co-$Z_{2^k}$-linearity means that the code can be\nobtained from a linear $Z_{2^k}$-code with the help of the new generalized Gray\nmap. Keywords: Gray map, Hadamard codes, MacWilliams identity, perfect codes,\n$Z_{2^k}$-linearity\n", " The class of poset metrics is very large and contains some interesting\nfamilies of metrics. A family of metrics, based on posets which are formed from\ndisjoint chains which have the same size, is examined. A necessary and\nsufficient condition, for the existence of perfect single-error-correcting\ncodes for such poset metrics, is proved.\n", " We study circle valued maps and consider the persistence of the homology of\ntheir fibers. The outcome is a finite collection of computable invariants which\nanswer the basic questions on persistence and in addition encode the topology\nof the source space and its relevant subspaces. Unlike persistence of real\nvalued maps, circle valued maps enjoy a different class of invariants called\nJordan cells in addition to bar codes. We establish a relation between the\nhomology of the source space and of its relevant subspaces with these\ninvariants and provide a new algorithm to compute these invariants from an\ninput matrix that encodes a circle valued map on an input simplicial complex.\n", " Let $\\mathcal{C}$ be a set of $m$ by $n$ matrices over $\\mathbb{F}_q$ such\nthat the rank of $A-B$ is at least $d$ for all distinct $A,B\\in \\mathcal{C}$.\nSuppose that $m\\leqslant n$. If $\\#\\mathcal{C}= q^{n(m-d+1)}$, then\n$\\mathcal{C}$ is a maximum rank distance (MRD for short) code. Until 2016,\nthere were only two known constructions of MRD codes for arbitrary $1<d<m-1$.\nOne was found by Delsarte (1978) and Gabidulin (1985) independently, and it was\nlater generalized by Kshevetskiy and Gabidulin (2005). We often call them\n(generalized) Gabidulin codes. Another family was recently obtained by Sheekey\n(2016), and its elements are called twisted Gabidulin codes. In the same paper,\nSheekey also proposed a generalization of the twisted Gabidulin codes. However\nthe equivalence problem for it is not considered, whence it is not clear\nwhether there exist new MRD codes in this generalization. We call the members\nof this putative larger family generalized twisted Gabidulin codes. In this\npaper, we first compute the Delsarte duals and adjoint codes of them, then we\ncompletely determine the equivalence between different generalized twisted\nGabidulin codes. In particular, it can be proven that, up to equivalence,\ngeneralized Gabidulin codes and twisted Gabidulin codes are both proper subsets\nof this family.\n", " In this paper, we introduce and study the Weyl transform of functions which\nare integrable with respect to a vector measure on a phase space associated to\na locally compact abelian group. We also study the Weyl transform of vector\nmeasures. Later, we also introduce and study the convolution of functions from\n$L^p$-spaces associated to a vector measure. We also study the Weyl transform\nof vector-valued functions and prove a vector-valued analogue of the\nHausdorff-Young inequality.\n", " Neural codes serve as a language for neurons in the brain. Convex codes,\nwhich arise from the pattern of intersections of convex sets in Euclidean\nspace, are of particular relevance to neuroscience. Not every code is convex,\nhowever, and the combinatorial properties of a code that determine its\nconvexity are still poorly understood. Here we find that a code that can be\nrealized by a collection of open convex sets may or may not be realizable by\nclosed convex sets, and vice versa, establishing that open convex and closed\nconvex codes are distinct classes. We also prove that max intersection-complete\ncodes (i.e. codes that contain all intersections of maximal codewords) are both\nopen convex and closed convex, and provide an upper bound for their minimal\nembedding dimension. Finally, we show that the addition of non-maximal\ncodewords to an open convex code preserves convexity.\n", " On a Goppa code whose structure polynomial has coefficients in the symbol\nfield, the Frobenius acts. Its fixed codewords form a subcode. Deleting the\nnaturally occurred redundance, we obtain a new code. It is proved that these\nnew codes approach the Gilbert-Varshamov bound. It is also proved that these\ncodes can be decoded within $O(n^2(\\logn)^a)$ operations in the symbol field,\nwhich is usually much small than the location field, where $n$ is the codeword\nlength, and $a$ a constant determined by the polynomial factorization\nalgorithm.\n", " In the present paper we introduce and study finite point subsets of a special\nkind, called optimum distributions, in the n-dimensional unit cube. Such\ndistributions are closely related with known (delta,s,n)-nets of low\ndiscrepancy. It turns out that optimum distributions have a rich combinatorial\nstructure. Namely, we show that optimum distributions can be characterized\ncompletely as maximum distance separable codes with respect to a non-Hamming\nmetric. Weight spectra of such codes can be evaluated precisely. We also\nconsider linear codes and distributions and study their general properties\nincluding the duality with respect to a suitable inner product. The\ncorresponding generalized MacWilliams identities for weight enumerators are\nbriefly discussed. Broad classes of linear maximum distance separable codes and\nlinear optimum distributions are explicitely constructed in the paper by the\nHermite interpolations over finite fields.\n", " Networks of neurons in the brain encode preferred patterns of neural activity\nvia their synaptic connections. Despite receiving considerable attention, the\nprecise relationship between network connectivity and encoded patterns is still\npoorly understood. Here we consider this problem for networks of\nthreshold-linear neurons whose computational function is to learn and store a\nset of binary patterns (e.g., a neural code) as \"permitted sets\" of the\nnetwork. We introduce a simple Encoding Rule that selectively turns \"on\"\nsynapses between neurons that co-appear in one or more patterns. The rule uses\nsynapses that are binary, in the sense of having only two states (\"on\" or\n\"off\"), but also heterogeneous, with weights drawn from an underlying synaptic\nstrength matrix S. Our main results precisely describe the stored patterns that\nresult from the Encoding Rule -- including unintended \"spurious\" states -- and\ngive an explicit characterization of the dependence on S. In particular, we\nfind that binary patterns are successfully stored in these networks when the\nexcitatory connections between neurons are geometrically balanced -- i.e., they\nsatisfy a set of geometric constraints. Furthermore, we find that certain types\nof neural codes are \"natural\" in the context of these networks, meaning that\nthe full code can be accurately learned from a highly undersampled set of\npatterns. Interestingly, many commonly observed neural codes in cortical and\nhippocampal areas are natural in this sense. As an application, we construct\nnetworks that encode hippocampal place field codes nearly exactly, following\npresentation of only a small fraction of patterns. To obtain our results, we\nprove new theorems using classical ideas from convex and distance geometry,\nsuch as Cayley-Menger determinants, revealing a novel connection between these\nareas of mathematics and coding properties of neural networks.\n", " The aim of this paper is to present some results about the space L^\\Phi(\\nu),\nwhere \\nu is a vector measure on a compact (not necessarily abelian) group and\n\\Phi is a Young function. We show that under certain conditions, the space\nL^\\Phi(\\nu) becomes an L^1(G)-module with respect to the usual convolution of\nfunctions. We also define one more convolution structure on L^\\Phi(\\nu).\n", " Property testing has been a major area of research in computer science in the\nlast three decades. By property testing we refer to an ensemble of problems,\nresults and algorithms which enable to deduce global information about some\ndata by only reading small random parts of it. In recent years, this theory\nfound its way into group theory, mainly via group stability.\n In this paper, we study the following problem: Devise a randomized algorithm\nthat given a subgroup $H$ of $G$, decides whether $H$ is the whole group or a\nproper subgroup, by checking whether a single (random) element of $G$ is in\n$H$. The search for such an algorithm boils down to the following purely group\ntheoretic problem: For $G$ of rank $k$, find a small as possible test subset\n$A\\subseteq G$ such that for every proper subgroup $H$, $|H\\cap A|\\leq\n(1-\\delta)|A|$ for some absolute constant $\\delta>0$, which we call the\ndetection probability of $A$. It turns out that the search for sets $A$ of size\nlinear in $k$ and constant detection probability is a non-commutative analogue\nof the classical search for families of good error correcting codes. This paper\nis devoted to proving that such test subsets exist, which implies good\nuniversal error correcting codes exist -- providing a far reaching\ngeneralization of the classical result of Shannon.\n In addition, we study this problem in certain subclasses of groups -- such as\nabelian, nilpotent, and finite solvable groups -- providing different\nconstructions of test subsets for these subclasses with various qualities.\nFinally, this generalized theory of non-commutative error correcting codes\nsuggests a plethora of interesting problems and research directions.\n", " The exploration of linear subspaces, particularly scattered subspaces, has\ngarnered considerable attention across diverse mathematical disciplines in\nrecent years, notably within finite geometries and coding theory. Scattered\nsubspaces play a pivotal role in analyzing various geometric structures such as\nblocking sets, two-intersection sets, complete arcs, caps in affine and\nprojective spaces over finite fields and rank metric codes. This paper\nintroduces a new infinite family of $h$-subspaces, along with their associated\nMRD codes. Additionally, it addresses the task of determining the generalized\nweights of these codes. Notably, we demonstrate that these MRD codes exhibit\nsome larger generalized weights compared to those previously identified.\n", " The divisible sandpile model is a fixed-energy continuous counterpart of the\nAbelian sandpile model. We start with a random initial configuration and\nredistribute mass deterministically. Under certain conditions the sandpile will\nstabilize. The associated odometer function describes the amount of mass\nemitted from each vertex during stabilization. In this survey we describe\nrecent scaling limit results of the odometer function depending on different\ninitial configurations and redistribution rules. Moreover we review connections\nto the obstacle problem from potential theory, including the connection between\nodometers and limiting shapes of growth models such as iDLA. Finally we state\nsome open problems.\n", " Over the past few decades, there has been extensive research on scattered\nsubspaces, partly because of their link to MRD codes. These subspaces can be\ncharacterized using linearized polynomials over finite fields. Within this\ncontext, scattered sequences extend the concept of scattered polynomials and\ncan be viewed as geometric equivalents of exceptional MRD codes. Up to now,\nonly scattered sequences of orders one and two have been developed. However,\nthis paper presents an infinite series of exceptional scattered sequences of\nany order beyond two which correspond to scattered subspaces that cannot be\nobtained as direct sum of scattered subspaces in smaller dimensions. The paper\nalso addresses equivalence concerns within this framework.\n", " In light of recent results from the LHC, MEG and neutrino experiments, we\nrevisit the issue of charged lepton flavour violation (LFV) in supersymmetric\ntheories with massive neutrinos, where flavour-violating soft\nsupersymmetry-breaking masses for sleptons are induced naturally by radiative\ncorrections. We link our results to the expectations for neutrino mixing angles\nin SU(5) enhanced by an abelian flavour symmetry, with particular focus on\n$\\theta_{13}$. We focus on the radiative decays $\\ell_i \\rightarrow \\ell_j\n\\gamma$ and on detection prospects at the LHC and a linear collider (LC). We\nuse supersymmetric parameters consistent with cosmological considerations and\nwith LHC searches for supersymmetry and the Higgs mass. We find a class of\nscenarios where the LHC may be sensitive to LFV sparticle decays and LFV\nprocesses could be detectable at a LC with centre-of-mass energy above 1 TeV,\nwhereas LFV lepton decays may be suppressed by cancellations in the decay\namplitudes.\n", " Large language models (LLMs) like Llama, Baichuan and Bloom models show\nremarkable ability with instruction fine-tuning in many natural language tasks.\nNevertheless, for the dialogue summarization task, which aims to generate\nsummaries for different roles in dialogue, most of the state-of-the-art methods\nconduct on small models (e.g Bart and Bert). Existing methods try to add task\nspecified optimization on small models like adding global-local centrality\nscore to models. In this paper, we propose an instruction fine-tuning model:\nBaichuan2-Sum, for role-oriented diaglouge summarization. By setting different\ninstructions for different roles, the model can learn from the dialogue\ninteractions and output the expected summaries. Furthermore, we applied NEFTune\ntechnique to add suitable noise during training to improve the results. The\nexperiments demonstrate that the proposed model achieves the new\nstate-of-the-art results on two public dialogue summarization datasets: CSDS\nand SAMSUM. We release our model and related codes to facilitate future studies\non dialogue summarization task.\n", " We discuss, in terms of rate-distortion theory, the fitness of molecular\ncodes as the problem of designing an optimal information channel. The fitness\nis governed by an interplay between the cost and quality of the channel, which\ninduces smoothness in the code. By incorporating this code fitness into\npopulation dynamics models, we suggest that the emergence and evolution of\nmolecular codes may be explained by simple channel design considerations.\n", " We consider a mobile edge computing scenario where users want to perform a\nlinear inference operation $\\boldsymbol{W} \\boldsymbol{x}$ on local data\n$\\boldsymbol{x}$ for some network-side matrix $\\boldsymbol{W}$. The inference\nis performed in a distributed fashion over multiple servers at the network\nedge. For this scenario, we propose a coding scheme that combines a rateless\ncode to provide resiliency against straggling servers--hence reducing the\ncomputation latency--and an irregular-repetition code to provide spatial\ndiversity--hence reducing the communication latency. We further derive a lower\nbound on the total latency--comprising computation latency, communication\nlatency, and decoding latency. The proposed scheme performs remarkably close to\nthe bound and yields significantly lower latency than the scheme based on\nmaximum distance separable codes recently proposed by Zhang and Simeone.\n", " This article studies codimension one foliations on projective man-ifolds\nhaving a compact leaf (free of singularities). It explores the interplay\nbetween Ueda theory (order of flatness of the normal bundle) and the holo-nomy\nrepresentation (dynamics of the foliation in the transverse direction). We\naddress in particular the following problems: existence of foliation having as\na leaf a given hypersurface with topologically torsion normal bundle, global\nstructure of foliations having a compact leaf whose holonomy is abelian (resp.\nsolvable), and factorization results.\n", " For a left action $S\\overset{\\lambda}{\\curvearrowright}X$ of a cancellative\nright amenable monoid $S$ on a discrete Abelian group $X$, we construct its Ore\nlocalization $G\\overset{\\lambda^*}{\\curvearrowright}X^*$, where $G$ is the\ngroup of left fractions of $S$; analogously, for a right action\n$K\\overset{\\rho}\\curvearrowleft S$ on a compact space $K$, we construct its Ore\ncolocalization $K^*\\overset{\\rho^*}{\\curvearrowleft} G$. Both constructions\npreserve entropy, i.e., for the algebraic entropy $h_{\\mathrm{alg}}$ and for\nthe topological entropy $h_{\\mathrm{top}}$ one has\n$h_{\\mathrm{alg}}(\\lambda)=h_{\\mathrm{alg}}(\\lambda^*)$ and\n$h_{\\mathrm{top}}(\\rho)=h_{\\mathrm{top}}(\\rho^*)$, respectively.\n Exploiting these constructions and the theory of quasi-tilings, we extend the\nAddition Theorem for $h_{\\mathrm{top}}$, known for right actions of countable\namenable groups on compact metrizable groups, to right actions\n$K\\overset{\\rho}{\\curvearrowleft} S$ of cancellative right amenable monoids $S$\n(with no restrictions on the cardinality) on arbitrary compact groups $K$.\n When the compact group $K$ is Abelian, we prove that $h_{\\mathrm{top}}(\\rho)$\ncoincides with $h_{\\mathrm{alg}}(\\hat{\\rho})$, where\n$S\\overset{\\hat{\\rho}}\\curvearrowright X$ is the dual left action on the\ndiscrete Pontryagin dual $X=\\hat{K}$, that is, a so-called Bridge Theorem. From\nthe Addition Theorem for $h_{\\mathrm{top}}$ and the Bridge Theorem, we obtain\nan Addition Theorem for $h_{\\mathrm{alg}}$ for left actions\n$S\\overset{\\lambda}\\curvearrowright X$ on discrete Abelian groups, so far known\nonly under the hypotheses that either $X$ is torsion or $S$ is locally\nmonotileable.\n The proofs substantially use the unified approach towards entropy based on\nthe entropy of actions of cancellative right amenable monoids on appropriately\ndefined normed monoids.\n", " We focus on quantum systems represented by a Hilbert space $L^2(A)$, where\n$A$ is a locally compact Abelian group that contains a compact open subgroup.\nWe examine two interconnected issues related to Weyl-Heisenberg operators.\nFirst, we provide a complete and elegant solution to the problem of describing\nthe stabilizer states in terms of their wave functions, an issue that arises in\nquantum information theory. Subsequently, we demonstrate that the stabilizer\nstates are precisely the minimizers of the Wehrl entropy functional, thereby\nresolving the analog of the Wehrl conjecture for any such group. Additionally,\nwe construct a moduli space for the set of stabilizer states, that is, a\nparameterization of this set, that endows it with a natural algebraic\nstructure, and we derive a formula for the number of stabilizer states when $A$\nis finite. Notably, these results are novel even for finite Abelian groups.\n", " We study questions in incidence geometry where the precise position of points\nis `blurry' (e.g. due to noise, inaccuracy or error). Thus lines are replaced\nby narrow tubes, and more generally affine subspaces are replaced by their\nsmall neighborhood. We show that the presence of a sufficiently large number of\napproximately collinear triples in a set of points in d dimensional complex\nspace implies that the points are close to a low dimensional affine subspace.\nThis can be viewed as a stable variant of the Sylvester-Gallai theorem and its\nextensions.\n Building on the recently found connection between Sylvester-Gallai type\ntheorems and complex Locally Correctable Codes (LCCs), we define the new notion\nof stable LCCs, in which the (local) correction procedure can also handle small\nperturbations in the euclidean metric. We prove that such stable codes with\nconstant query complexity do not exist. No impossibility results were known in\nany such local setting for more than 2 queries.\n", " Let $(G, +)$ be an abelian group. In 2004, Eliahou and Kervaire found an\nexplicit formula for the smallest possible cardinality of the sumset $A+A$,\nwhere $A \\subseteq G$ has fixed cardinality $r$. We consider instead the\nsmallest possible cardinality of the difference set $A-A$, which is always\ngreater than or equal to the smallest possible cardinality of $A+A$ and can be\nstrictly greater. We conjecture a formula for this quantity and prove the\nconjecture in the case that $G$ is a cyclic group or a vector space over a\nfinite field. This resolves a conjecture of Bajnok and Matzke on signed\nsumsets.\n", " The use of graph centrality measures applied to biological networks, such as\nprotein interaction networks, underpins much research into identifying key\nplayers within biological processes. This approach however is restricted to\ndyadic interactions and it is well-known that in many instances interactions\nare polyadic. In this study we illustrate the merit of using hypergraph\ncentrality applied to a hypernetwork as an alternative. Specifically, we review\nand propose an extension to a recently introduced node and edge nonlinear\nhypergraph centrality model which provides mutually dependent node and edge\ncentralities. A Saccharomyces Cerevisiae protein complex hypernetwork is used\nas an example application with nodes representing proteins and hyperedges\nrepresenting protein complexes. The resulting rankings of the nodes and edges\nare considered to see if they provide insight into the essentiality of the\nproteins and complexes. We find that certain variations of the model predict\nessentiality more accurately and that the degree-based variation illustrates\nthat the centrality-lethality rule extends to a hypergraph setting. In\nparticular, through exploitation of the models flexibility, we identify small\nsets of proteins densely populated with essential proteins.\n", " Using the concept of projective systems for linear codes and elementary\nlinear algebra, we show that projective $[n,k]_q$ codes form a connected\nsubgraph in the Grassmann graph consisting of $k$-dimensional subspaces of an\n$n$-dimensional vector space over the $q$-element field.\n", " To address growth challenges facing large Data Centers and supercomputing\nclusters a new construction is presented for scalable, high throughput, low\nlatency networks. The resulting networks require 1.5-5 times fewer switches,\n2-6 times fewer cables, have 1.2-2 times lower latency and correspondingly\nlower congestion and packet losses than the best present or proposed networks\nproviding the same number of ports at the same total bisection. These advantage\nratios increase with network size. The key new ingredient is the exact\nequivalence discovered between the problem of maximizing network bisection for\nlarge classes of practically interesting Cayley graphs and the problem of\nmaximizing codeword distance for linear error correcting codes. Resulting\ntranslation recipe converts existent optimal error correcting codes into\noptimal throughput networks.\n", " An alternative proof is given of the existence of greatest lower bounds in\nthe imbalance order of binary maximal instantaneous codes of a given size.\nThese codes are viewed as maximal antichains of a given size in the infinite\nbinary tree of 0-1 words. The proof proposed makes use of a single balancing\noperation instead of expansion and contraction as in the original proof of the\nexistence of glb.\n", " Grid computing is distributed computing performed transparently across\nmultiple administrative domains. Grid middleware, which is meant to enable\naccess to grid resources, is currently widely seen as being too heavyweight\nand, in consequence, unwieldy for general scientific use. Its heavyweight\nnature, especially on the client-side, has severely restricted the uptake of\ngrid technology by computational scientists. In this paper, we describe the\nApplication Hosting Environment (AHE) which we have developed to address some\nof these problems. The AHE is a lightweight, easily deployable environment\ndesigned to allow the scientist to quickly and easily run legacy applications\non distributed grid resources. It provides a higher level abstraction of a grid\nthan is offered by existing grid middleware schemes such as the Globus Toolkit.\nAs a result the computational scientist does not need to know the details of\nany particular underlying grid middleware and is isolated from any changes to\nit on the distributed resources. The functionality provided by the AHE is\n`application-centric': applications are exposed as web services with a\nwell-defined standards-compliant interface. This allows the computational\nscientist to start and manage application instances on a grid in a transparent\nmanner, thus greatly simplifying the user experience. We describe how a range\nof computational science codes have been hosted within the AHE and how the\ndesign of the AHE allows us to implement complex workflows for deployment on\ngrid infrastructure.\n", " We propose an Abelian gauged version of the singlet-doublet fermionic dark\nmatter (DM) model where DM, combination of a vector like fermion doublet and a\nfermion singlet, is naturally stabilised by the gauge symmetry without\nrequiring any ad-hoc discrete symmetries. In order to have good detection\nprospects at collider experiments like the large hadron collider (LHC) and\nenlarged parameter space for low mass DM, we consider the additional gauge\nsymmetry to be based on the quantum $B-3L_{\\tau}$ where the restriction to\nthird generation of leptons is chosen to have weaker bounds from the LHC on the\ncorresponding gauge boson. The triangle anomalies arising in this model can be\ncancelled by including a right handed neutrino which takes part in generating\nlight neutrino masses through type I seesaw mechanism. Apart from DM, collider\nprospects and light neutrino masses, the model also offers high scale validity\ngiving rise to a stable electroweak vacuum and perturbative couplings all the\nway up to the Planck scale. We constrain our model parameters from these\nrequirements as well as existing relevant constraints related to DM and\ncolliders.\n", " After 10 years of steadily increasing the experimental precision at LEP/SLC,\nthere is a strong demand on an update of existing programs for fermion pair\nproduction. We present a rederivation of the O(alpha) Bremsstrahlung\ncorrections to e+e- --> f+f- for the semi-analytic program ZFITTER. We focus on\nobservables like total cross section and forward-backward asymmetry in the\nleptonic case with combined cuts on acollinearity angle, acceptance angle, and\nminimal energy of the fermions. The outcome of our analysis is a shift of the\npredictions by ZFITTER at LEP 1 energies off-resonance of a few per mil while\nat the Z resonance numerical changes can be neglected. Thus we obtain for cross\nsections and asymmetries at LEP 1 a level of agreement with other programs of\nbetter than per mil, like for the kinematically simpler s' cut option. A\npreliminary analysis of ZFITTER, TOPAZ0, and other codes at LEP 2 energies\nshowing deviations of several per cent with acollinearity cuts enforce a future\nexamination of higher order effects with different cuts. The predictions by\nLEP/SLC data, however, are not affected within the experimental errors.\n", " Motivated by the need to correct the potentially large kinematic errors in\napproximations used in the standard formulation of perturbative QCD, we\nreformulate deeply inelastic lepton-proton scattering in terms of gauge\ninvariant, universal parton correlation functions which depend on all\ncomponents of parton four-momentum. Currently, different hard QCD processes are\ndescribed by very different perturbative formalisms, each relying on its own\nset of kinematical approximations. In this paper we show how to set up\nformalism that avoids approximations on final-state momenta, and thus has a\nvery general domain of applicability. The use of exact kinematics introduces a\nnumber of significant conceptual shifts already at leading order, and tightly\nconstrains the formalism. We show how to define parton correlation functions\nthat generalize the concepts of parton density, fragmentation function, and\nsoft factor. After setting up a general subtraction formalism, we obtain a\nfactorization theorem. To avoid complications with Ward identities the full\nderivation is restricted to abelian gauge theories; even so the resulting\nstructure is highly suggestive of a similar treatment for non-abelian gauge\ntheories.\n", " Consider a locally compact quantum group $\\mathbb{G}$ with a closed classical\nabelian subgroup $\\Gamma$ equipped with a $2$-cocycle\n$\\Psi:\\hat{\\Gamma}\\times\\hat{\\Gamma}\\to\\mathbb{C}$. We study in detail the\nassociated Rieffel deformation $\\mathbb{G}^{\\Psi}$ and establish a canonical\ncorrespondence between $\\Gamma$-invariant convolution semigroups of states on\n$\\mathbb{G}$ and on $\\mathbb{G}^{\\Psi}$.\n", " Geometric disentanglement, the separation of latent codes for intrinsic (i.e.\nidentity) and extrinsic(i.e. pose) geometry, is a prominent task for generative\nmodels of non-Euclidean data such as 3D deformable models. It provides greater\ninterpretability of the latent space, and leads to more control in generation.\nThis work introduces a mesh feature, the conformal factor and normal feature\n(CFAN),for use in mesh convolutional autoencoders. We further propose CFAN-VAE,\na novel architecture that disentangles identity and pose using the CFAN\nfeature. Requiring no label information on the identity or pose during\ntraining, CFAN-VAE achieves geometric disentanglement in an unsupervisedway.\nOur comprehensive experiments, including reconstruction, interpolation,\ngeneration, and identity/pose transfer, demonstrate CFAN-VAE achieves\nstate-of-the-art performance on unsupervised geometric disentanglement. We also\nsuccessfully detect a level of geometric disentanglement in mesh convolutional\nautoencoders that encode xyz-coordinates directly by registering its latent\nspace to that of CFAN-VAE.\n", " In this paper, we introduce the concept of the complete cycle index and\ndiscuss a relation with the complete weight enumerator in coding theory. This\nwork was motivated by Cameron's lecture note \"Polynomial aspects of codes,\nmatroids and permutation groups.\"\n" ]

[ " For strongly convex objectives that are smooth, the classical theory of\ngradient descent ensures linear convergence relative to the number of gradient\nevaluations. An analogous nonsmooth theory is challenging. Even when the\nobjective is smooth at every iterate, the corresponding local models are\nunstable and the number of cutting planes invoked by traditional remedies is\ndifficult to bound, leading to convergences guarantees that are sublinear\nrelative to the cumulative number of gradient evaluations. We instead propose a\nmultipoint generalization of the gradient descent iteration for local\noptimization. While designed with general objectives in mind, we are motivated\nby a ``max-of-smooth'' model that captures the subdifferential dimension at\noptimality. We prove linear convergence when the objective is itself\nmax-of-smooth, and experiments suggest a more general phenomenon.\n", " One of the most important parts of Artificial Neural Networks is minimizing\nthe loss functions which tells us how good or bad our model is. To minimize\nthese losses we need to tune the weights and biases. Also to calculate the\nminimum value of a function we need gradient. And to update our weights we need\ngradient descent. But there are some problems with regular gradient descent ie.\nit is quite slow and not that accurate. This article aims to give an\nintroduction to optimization strategies to gradient descent. In addition, we\nshall also discuss the architecture of these algorithms and further\noptimization of Neural Networks in general\n", " Stochastic gradient descent (SGD) is a powerful method for large-scale\noptimization problems in the area of machine learning, especially for a\nfinite-sum formulation with numerous variables. In recent years, mini-batch SGD\ngains great success and has become a standard technique for training deep\nneural networks fed with big amount of data. Inspired by its success in deep\nlearning, we apply the idea of SGD with batch selection of samples to a classic\noptimization problem in decision version. Given $n$ unit circles, the equal\ncircle packing problem (ECPP) asks whether there exist a feasible packing that\ncould put all the circles inside a circular container without overlapping.\nSpecifically, we propose a stochastic item descent method (SIDM) for ECPP in\nlarge scale, which randomly divides the unit circles into batches and runs\nBroyden-Fletcher-Goldfarb-Shanno (BFGS) algorithm on the corresponding batch\nfunction iteratively to speedup the calculation. We also increase the batch\nsize during the batch iterations to gain higher quality solution. Comparing to\nthe current best packing algorithms, SIDM greatly speeds up the calculation of\noptimization process and guarantees the solution quality for large scale\ninstances with up to 1500 circle items, while the baseline algorithms usually\nhandle about 300 circle items. The results indicate the highly efficiency of\nSIDM for this classic optimization problem in large scale, and show potential\nfor other large scale classic optimization problems in which gradient descent\nis used for optimization.\n", " A candidate explanation of the good empirical performance of deep neural\nnetworks is the implicit regularization effect of first order optimization\nmethods. Inspired by this, we prove a convergence theorem for nonconvex\ncomposite optimization, and apply it to a general learning problem covering\nmany machine learning applications, including supervised learning. We then\npresent a deep multilayer perceptron model and prove that, when sufficiently\nwide, it $(i)$ leads to the convergence of gradient descent to a global optimum\nwith a linear rate, $(ii)$ benefits from the implicit regularization effect of\ngradient descent, $(iii)$ is subject to novel bounds on the generalization\nerror, $(iv)$ exhibits the lazy training phenomenon and $(v)$ enjoys learning\nrate transfer across different widths. The corresponding coefficients, such as\nthe convergence rate, improve as width is further increased, and depend on the\neven order moments of the data generating distribution up to an order depending\non the number of layers. The only non-mild assumption we make is the\nconcentration of the smallest eigenvalue of the neural tangent kernel at\ninitialization away from zero, which has been shown to hold for a number of\nless general models in contemporary works. We present empirical evidence\nsupporting this assumption as well as our theoretical claims.\n", " This paper is concerned with the utilization of deterministically modeled\nchemical reaction networks for the implementation of (feed-forward) neural\nnetworks. We develop a general mathematical framework and prove that the\nordinary differential equations (ODEs) associated with certain reaction network\nimplementations of neural networks have desirable properties including (i)\nexistence of unique positive fixed points that are smooth in the parameters of\nthe model (necessary for gradient descent), and (ii) fast convergence to the\nfixed point regardless of initial condition (necessary for efficient\nimplementation). We do so by first making a connection between neural networks\nand fixed points for systems of ODEs, and then by constructing reaction\nnetworks with the correct associated set of ODEs. We demonstrate the theory by\nconstructing a reaction network that implements a neural network with a\nsmoothed ReLU activation function, though we also demonstrate how to generalize\nthe construction to allow for other activation functions (each with the\ndesirable properties listed previously). As there are multiple types of\n\"networks\" utilized in this paper, we also give a careful introduction to both\nreaction networks and neural networks, in order to disambiguate the overlapping\nvocabulary in the two settings and to clearly highlight the role of each\nnetwork's properties.\n", " We use the framework of the first-order differential structure in metric\nmeasure spaces introduced by Gigli to define a notion of weak solutions to\ngradient flows of convex, lower semicontinuous and coercive functionals. We\nprove their existence and uniqueness and show that they are also variational\nsolutions; in particular, this is an existence result for variational\nsolutions. Then, we apply this technique in the case of a gradient flow of a\nfunctional with inhomogeneous growth.\n", " We derive fundamental lower bounds on the connectivity and the memory\nrequirements of deep neural networks guaranteeing uniform approximation rates\nfor arbitrary function classes in $L^2(\\mathbb R^d)$. In other words, we\nestablish a connection between the complexity of a function class and the\ncomplexity of deep neural networks approximating functions from this class to\nwithin a prescribed accuracy. Additionally, we prove that our lower bounds are\nachievable for a broad family of function classes. Specifically, all function\nclasses that are optimally approximated by a general class of representation\nsystems---so-called \\emph{affine systems}---can be approximated by deep neural\nnetworks with minimal connectivity and memory requirements. Affine systems\nencompass a wealth of representation systems from applied harmonic analysis\nsuch as wavelets, ridgelets, curvelets, shearlets, $\\alpha$-shearlets, and more\ngenerally $\\alpha$-molecules. Our central result elucidates a remarkable\nuniversality property of neural networks and shows that they achieve the\noptimum approximation properties of all affine systems combined. As a specific\nexample, we consider the class of $\\alpha^{-1}$-cartoon-like functions, which\nis approximated optimally by $\\alpha$-shearlets. We also explain how our\nresults can be extended to the case of functions on low-dimensional immersed\nmanifolds. Finally, we present numerical experiments demonstrating that the\nstandard stochastic gradient descent algorithm generates deep neural networks\nproviding close-to-optimal approximation rates. Moreover, these results\nindicate that stochastic gradient descent can actually learn approximations\nthat are sparse in the representation systems optimally sparsifying the\nfunction class the network is trained on.\n", " Adam is a popular variant of stochastic gradient descent for finding a local\nminimizer of a function. In the constant stepsize regime, assuming that the\nobjective function is differentiable and non-convex, we establish the\nconvergence in the long run of the iterates to a stationary point under a\nstability condition. The key ingredient is the introduction of a\ncontinuous-time version of Adam, under the form of a non-autonomous ordinary\ndifferential equation. This continuous-time system is a relevant approximation\nof the Adam iterates, in the sense that the interpolated Adam process converges\nweakly towards the solution to the ODE. The existence and the uniqueness of the\nsolution are established. We further show the convergence of the solution\ntowards the critical points of the objective function and quantify its\nconvergence rate under a Lojasiewicz assumption. Then, we introduce a novel\ndecreasing stepsize version of Adam. Under mild assumptions, it is shown that\nthe iterates are almost surely bounded and converge almost surely to critical\npoints of the objective function. Finally, we analyze the fluctuations of the\nalgorithm by means of a conditional central limit theorem.\n", " This paper identifies certain interesting mathematical problems of stochastic\nquantization type in the modeling of Laser propagation through turbulent media.\nIn some of the typical physical contexts the problem reduces to stochastic\nSchrodinger equation with space-time white noise of Gaussian, Poisson and Levy\ntype. We identify their mathematical resolution via stochastic quantization.\nNonlinear phenomena such as Kerr effect can be modeled by stochastic nonlinear\nSchrodinger equation in the focusing case with space-time white noise. A\ntreatment of stochastic transport equation, the Korteweg-de Vries Equation as\nwell as a number of other nonlinear wave equations with space-time white noise\nis also given. Main technique is the S-transform (we will actually use closely\nrelated Hermite transform) which converts the stochastic partial differential\nequation with space time white noise to a deterministic partial differential\nequation defined on the Hida-Kondratiev white noise distribution space. We then\nutlize the inverse S-transform/Hermite transform known as the characterization\ntheorem combined with the infinite dimensional implicit function theorem for\nanalytic maps to establish local existence and uniqueness theorems for pathwise\nsolutions of these class of problems. The particular focus of this paper on\nsingular white noise distributions is motivated by practical situations where\nthe refractive index fluctuations in propagation medium in space and time are\nintense due to turbulence, ionospheric plasma turbulence, marine-layer\nfluctuations, etc. Since a large class of partial differential equations that\narise in nonlinear wave propagation have polynomial type nonlinearities, white\nnoise distribution theory is an effective tool in studying these problems\nsubject to different types of white noises.\n", " We study the Helton-Howe trace and the Connes-Chern character for Toeplitz\noperators on weighted Bergman spaces via the idea of quantization. We prove a\nlocal formula for the large $t$-limit of the Connes-Chern character as the\nweight goes to infinity. And we show that the Helton-Howe trace of Toeplitz\noperators is independent of the weight $t$ and obtain a local formula for the\nHelton-Howe trace for all weighted Bergman spaces using harmonic analysis and\nquantization.\n", " Although Ornstein's nonestimate entails the impossibility to control in\ngeneral all the $L^1$-norm of derivatives of a function by the $L^1$-norm of a\nconstant coefficient homogeneous vector differential operator, the\ncorresponding endpoint Sobolev inequality has been known to hold in many cases:\nthe gradient of scalar functions (Gagliardo and Nirenberg), the deformation\noperator (Korn-Sobolev inequality by M.J. Strauss), and the Hodge complex\n(Bourgain and Brezis). The class of differential operators for which estimates\nholds can be characterized by a cancelling condition. The proof of the\nestimates rely on a duality estimate for $L^1$-vector fields lying in the\nkernel of a cocancelling differential operator, combined with classical linear\nalgebra and harmonic analysis techniques. This characterization unifies classes\nof known Sobolev inequalities and extends to fractional Sobolev and Hardy\ninequalities. A similar weaker condition introduced by Rai\\c{t}\\u{a}\ncharacterizes the operators for which there is an $L^\\infty$-estimate on\nlower-order derivatives.\n", " We define an integral, the distributional integral of functions of one real\nvariable, that is more general than the Lebesgue and the\nDenjoy-Perron-Henstock-Kurzweil integrals, and which allows the integration of\nfunctions with distributional values everywhere or nearly everywhere.\n Our integral has the property that if $f$ is locally distributionally\nintegrable over the real line and $\\psi\\in\\mathcal{D}(\\mathbb{R}%) $ is a test\nfunction, then $f\\psi$ is distributionally integrable, and the formula%\n[<\\mathsf{f},\\psi> =(\\mathfrak{dist}) \\int_{-\\infty}^{\\infty}f(x) \\psi(x)\n\\,\\mathrm{d}% x\\,,] defines a distribution\n$\\mathsf{f}\\in\\mathcal{D}^{\\prime}(\\mathbb{R}) $ that has distributional point\nvalues almost everywhere and actually $\\mathsf{f}(x) =f(x) $ almost everywhere.\n The indefinite distributional integral $F(x) =(\\mathfrak{dist})\n\\int_{a}^{x}f(t) \\,\\mathrm{d}t$ corresponds to a distribution with point values\neverywhere and whose distributional derivative has point values almost\neverywhere equal to $f(x).$\n The distributional integral is more general than the standard integrals, but\nit still has many of the useful properties of those standard ones, including\nintegration by parts formulas, substitution formulas, even for infinite\nintervals --in the Ces\\`{a}ro sense--, mean value theorems, and convergence\ntheorems. The distributional integral satisfies a version of Hake's theorem.\nUnlike general distributions, locally distributionally integrable functions can\nbe restricted to closed sets and can be multiplied by power functions with real\npositive exponents.\n", " Real-valued functions on geometric data -- such as node attributes on a graph\n-- can be optimized using descriptors from persistent homology, allowing the\nuser to incorporate topological terms in the loss function. When optimizing a\nsingle real-valued function (the one-parameter setting), there is a canonical\nchoice of descriptor for persistent homology: the barcode. The operation\nmapping a real-valued function to its barcode is differentiable almost\neverywhere, and the convergence of gradient descent for losses using barcodes\nis relatively well understood. When optimizing a vector-valued function (the\nmultiparameter setting), there is no unique choice of descriptor for\nmultiparameter persistent homology, and many distinct descriptors have been\nproposed. This calls for the development of a general framework for\ndifferentiability and optimization that applies to a wide range of\nmultiparameter homological descriptors. In this article, we develop such a\nframework and show that it encompasses well-known descriptors of different\nflavors, such as signed barcodes and the multiparameter persistence landscape.\nWe complement the theory with numerical experiments supporting the idea that\noptimizing multiparameter homological descriptors can lead to improved\nperformances compared to optimizing one-parameter descriptors, even when using\nthe simplest and most efficiently computable multiparameter descriptors.\n", " For general compact K\\\"ahler manifolds it is shown that both Toeplitz\nquantization and geometric quantization lead to a well-defined (by operator\nnorm estimates) classical limit. This generalizes earlier results of the\nauthors and Klimek and Lesniewski obtained for the torus and higher genus\nRiemann surfaces, respectively. We thereby arrive at an approximation of the\nPoisson algebra by a sequence of finite-dimensional matrix algebras $gl(N)$,\n$N\\to\\infty$.\n", " We show that the Template Overlap Method can improve the signal to background\nratio of boosted $H\\to b \\bar b$ events produced in association with a\nleptonically decaying $W$. We introduce several improvements on the previous\nformulations of the template method. Varying three-particle template subcones\nincreases the rejection power against the backgrounds, while sequential\ntemplate generation ensures an efficient coverage in template phase space. We\nintegrate b-tagging information into the template overlap framework and\nintroduce a new template based observable, the template stretch. Our analysis\ntakes into account the contamination from the charm daughters of top decays in\n$t\\bar t$ events, and includes nearly-realistic effects of pileup and\nunderlying events. We show that the Template Overlap Method displays very low\nsensitivity to pileup, hence providing a self-contained alternative to other\nmethods of pile up subtraction. The developments described in this work are\nquite general, and may apply to other searches for massive boosted objects.\n", " We establish the various properties as well as diverse relations of the\nascent and descent spectra for bounded linear operators. We specially focus on\nthe theory of subspectrum. Furthermore, we construct a new concept of\nconvergence for such spectra.\n", " At the LHC, a generic supersymmetric cascade can be a source of top quark.\nSpecifically third generation squarks and gluino are the major sources of top\nquark which could also be boosted. In this article, we have shown that jet\nsubstructure algorithm can be very useful in identifying such boosted top\nquarks in the cascade. We take inclusive three jets plus zero lepton plus\nmissing energy final state and try to reconstruct at least one hadronically\ndecaying top quark by using top tagging techniques which has good prospect at\nthe LHC.\n", " We introduce a notion of {\\em cyclic Schur-positivity} for sets of\npermutations, which naturally extends the classical notion of Schur-positivity,\nand it involves the existence of a bijection from permutations to standard\nYoung tableaux that preserves the cyclic descent set. Cyclic Schur-positive\nsets of permutations are always Schur-positive, but the converse does not hold,\nas exemplified by inverse descent classes, Knuth classes and conjugacy classes.\n In this paper we show that certain classes of permutations invariant under\neither horizontal or vertical rotation are cyclic Schur-positive. The proof\nunveils a new equidistribution phenomenon of descent sets on permutations,\nprovides affirmative solutions to conjectures from [9] and [2], and yields new\nexamples of Schur-positive sets.\n", " We develop an efficient algorithm, coordinate descent FCI (CDFCI), for the\nelectronic structure ground state calculation in the configuration interaction\nframework. CDFCI solves an unconstrained non-convex optimization problem, which\nis a reformulation of the full configuration interaction eigenvalue problem,\nvia an adaptive coordinate descent method with a deterministic compression\nstrategy. CDFCI captures and updates appreciative determinants with different\nfrequencies proportional to their importance. We show that CDFCI produces\naccurate variational energy for both static and dynamic correlation by\nbenchmarking the binding curve of nitrogen dimer in the cc-pVDZ basis with\n$10^{-3}$ mHa accuracy. We also demonstrate the efficiency and accuracy of\nCDFCI for strongly correlated chromium dimer in the Ahlrichs VDZ basis and\nproduces state-of-the-art variational energy.\n", " An inconsistency of quantum field theory, regarding the signs of vacuum\nenergy and vacuum pressure of elementary fields versus non-elementary fields\n(like e.g. phonon fields), is pointed out. An improved law for the canonical\nquantization of fields is presented, which is based on the distinction between\nelementary and non-elementary fields. Remarkably, the improved quantization\nmethod removes not only the inconsistency of quantum field theory, but at the\nsame time solves the (old) cosmological constant problem for all fields of the\nstandard model of elementary particles (but not for the hypothetical inflaton\nfields), without compromising any of the achievements of established quantum\nfield theory.\n", " Two notions of riffle shuffling on finite Coxeter groups are given: one using\nSolomon's descent algebra and another using random walk on chambers of\nhyperplane arrangements. These coincide for types $A$,$B$,$C$, $H_3$, and rank\ntwo groups. Both notions have the same, simple eigenvalues. The hyperplane\ndefinition is especially natural and satisfies a positivity property when $W$\nis crystallographic and the relevant parameter is a good prime. The hyperplane\nviewpoint suggests interesting connections with Lie theory and leads to a\nnotion of riffle shuffling for arbitrary real hyperplane arrangements and\noriented matroids. Connections with Cellini's descent algebra are given.\n", " According to the wave power rule, the second derivative of a function with\nrespect to the variable t is equal to negative n times the function raised to\nthe power of 2n-1. Solving the ordinary differential equations numerically\nresults in waves appearing in the figures. The ordinary differential equation\nis very simple; however, waves, including the regular amplitude and period, are\ndrawn in the figure. In this study, the function for obtaining the wave is\ncalled the leaf function. Based on the leaf function, the exact solutions for\nthe undamped and unforced Duffing equations are presented. In the ordinary\ndifferential equation, in the positive region of the variable, the second\nderivative becomes negative. Therefore, in the case that the curves vary with\nthe time under the condition x(t)>0, the gradient constantly decreases as time\nincreases. That is, the tangential vector on the curve of the graph changes\nfrom the upper right direction to the lower right direction as time increases.\nOn the other hand, in the negative region of the variable, the second\nderivative becomes positive. The gradient constantly increases as time\ndecreases. That is, the tangent vector on the curve changes from the lower\nright direction to the upper right direction as time increases. Since the\nbehavior occurring in the positive region of the variable and the behavior\noccurring in the negative region of the variable alternately occur in regular\nintervals, waves appear by these interactions. In this paper, I present seven\ntypes of damped and divergence exact solutions by combining trigonometric\nfunctions, hyperbolic functions, hyperbolic leaf functions, leaf functions, and\nexponential functions. In each type, I show the derivation method and numerical\nexamples, as well as describe the features of the waveform.\n", " We study nonlinear degenerate parabolic equations of Fokker-Planck type which\ncan be viewed as gradient flows with respect to the recently introduced\nspherical Hellinger-Kantorovich distance. The driving entropy is not assumed to\nbe geodesically convex. We prove solvability of the problem and the\nentropy-entropy production inequality, which implies exponential convergence to\nthe equilibrium. As a corollary, we obtain some related results for the\nWasserstein gradient flows. We also deduce transportation inequalities in the\nspirit of Talagrand, Otto and Villani for the spherical and conic\nHellinger-Kantorovich distances.\n", " Bandstructure effects in the electronic transport of strongly quantized\nsilicon nanowire field-effect-transistors (FET) in various transport\norientations are examined. A 10-band sp3d5s* semi-empirical atomistic\ntight-binding model coupled to a self consistent Poisson solver is used for the\ndispersion calculation. A semi-classical, ballistic FET model is used to\nevaluate the current-voltage characteristics. It is found that the total gate\ncapacitance is degraded from the oxide capacitance value by 30% for wires in\nall the considered transport orientations ([100], [110], [111]). Different wire\ndirections primarily influence the carrier velocities, which mainly determine\nthe relative performance differences, while the total charge difference is\nweakly affected. The velocities depend on the effective mass and degeneracy of\nthe dispersions. The [110] and secondly the [100] oriented 3nm thick nanowires\nexamined, indicate the best ON-current performance compared to [111] wires. The\ndispersion features are strong functions of quantization. Effects such as\nvalley splitting can lift the degeneracies especially for wires with cross\nsection sides below 3nm. The effective masses also change significantly with\nquantization, and change differently for different transport orientations. For\nthe cases of [100] and [111] wires the masses increase with quantization,\nhowever, in the [110] case, the mass decreases. The mass variations can be\nexplained from the non-parabolicities and anisotropies that reside in the first\nBrillouin zone of silicon.\n", " In this survey, we review asymptotic results of some orthogonal polynomials\nwhich are relate to the Painlev\\'e transcendents. They are obtained by applying\nDeift-Zhou's nonlinear steepest descent method for Riemann-Hilbert problems. In\nthe last part of this article, we list several open problems.\n", " Chemical reactions occur in energy, environmental, biological, and many other\nnatural systems, and the inference of the reaction networks is essential to\nunderstand and design the chemical processes in engineering and life sciences.\nYet, revealing the reaction pathways for complex systems and processes is still\nchallenging due to the lack of knowledge of the involved species and reactions.\nHere, we present a neural network approach that autonomously discovers reaction\npathways from the time-resolved species concentration data. The proposed\nChemical Reaction Neural Network (CRNN), by design, satisfies the fundamental\nphysics laws, including the Law of Mass Action and the Arrhenius Law.\nConsequently, the CRNN is physically interpretable such that the reaction\npathways can be interpreted, and the kinetic parameters can be quantified\nsimultaneously from the weights of the neural network. The inference of the\nchemical pathways is accomplished by training the CRNN with species\nconcentration data via stochastic gradient descent. We demonstrate the\nsuccessful implementations and the robustness of the approach in elucidating\nthe chemical reaction pathways of several chemical engineering and biochemical\nsystems. The autonomous inference by the CRNN approach precludes the need for\nexpert knowledge in proposing candidate networks and addresses the curse of\ndimensionality in complex systems. The physical interpretability also makes the\nCRNN capable of not only fitting the data for a given system but also\ndeveloping knowledge of unknown pathways that could be generalized to similar\nchemical systems.\n", " Sparseness is a useful regularizer for learning in a wide range of\napplications, in particular in neural networks. This paper proposes a model\ntargeted at classification tasks, where sparse activity and sparse connectivity\nare used to enhance classification capabilities. The tool for achieving this is\na sparseness-enforcing projection operator which finds the closest vector with\na pre-defined sparseness for any given vector. In the theoretical part of this\npaper, a comprehensive theory for such a projection is developed. In\nconclusion, it is shown that the projection is differentiable almost everywhere\nand can thus be implemented as a smooth neuronal transfer function. The entire\nmodel can hence be tuned end-to-end using gradient-based methods. Experiments\non the MNIST database of handwritten digits show that classification\nperformance can be boosted by sparse activity or sparse connectivity. With a\ncombination of both, performance can be significantly better compared to\nclassical non-sparse approaches.\n", " Let $n$ be a nonnegative integer and $I$ be a finite set of positive\nintegers. In 1915, MacMahon proved that the number of permutations in the\nsymmetric group $\\mathfrak{S}_n$ with descent set $I$ is a polynomial in $n$.\nWe call this the descent polynomial. However, basic properties of these\npolynomials such as a description of their coefficients and roots do not seem\nto have been studied in the literature. Much more recently, in 2013, Billey,\nBurdzy, and Sagan showed that the number of elements of $\\mathfrak{S}_n$ with\npeak set $I$ is a polynomial in $n$ times a certain power of two. Since then,\nthere have been a flurry of papers investigating properties of this peak\npolynomial. The purpose of the present paper is to study the descent\npolynomial. We will see that it displays some interesting parallels with its\npeak relative. Conjectures and questions for future research are scattered\nthroughout.\n", " We demonstrate that the classification of boosted, hadronically-decaying weak\ngauge bosons can be significantly improved over traditional cut-based and\nBDT-based methods using deep learning and the jet charge variable. We construct\nbinary taggers for $W^+$ vs. $W^-$ and $Z$ vs. $W$ discrimination, as well as\nan overall ternary classifier for $W^+$/$W^-$/$Z$ discrimination. Besides a\nsimple convolutional neural network (CNN), we also explore a composite of two\nCNNs, with different numbers of layers in the jet $p_{T}$ and jet charge\nchannels. We find that this novel structure boosts the performance particularly\nwhen considering the $Z$ boson as signal. The methods presented here can\nenhance the physics potential in SM measurements and searches for new physics\nthat are sensitive to the electric charge of weak gauge bosons.\n", " Bilevel Optimization Programming is used to model complex and conflicting\ninteractions between agents, for example in Robust AI or Privacy-preserving AI.\nIntegrating bilevel mathematical programming within deep learning is thus an\nessential objective for the Machine Learning community. Previously proposed\napproaches only consider single-level programming. In this paper, we extend\nexisting single-level optimization programming approaches and thus propose\nDifferentiating through Bilevel Optimization Programming (BiGrad) for\nend-to-end learning of models that use Bilevel Programming as a layer. BiGrad\nhas wide applicability and can be used in modern machine learning frameworks.\nBiGrad is applicable to both continuous and combinatorial Bilevel optimization\nproblems. We describe a class of gradient estimators for the combinatorial case\nwhich reduces the requirements in terms of computation complexity; for the case\nof the continuous variable, the gradient computation takes advantage of the\npush-back approach (i.e. vector-jacobian product) for an efficient\nimplementation. Experiments show that the BiGrad successfully extends existing\nsingle-level approaches to Bilevel Programming.\n", " Online learning algorithms have become a ubiquitous tool in the machine\nlearning toolbox and are frequently used in small, resource-constraint\nenvironments. Among the most successful online learning methods are Decision\nTree (DT) ensembles. DT ensembles provide excellent performance while adapting\nto changes in the data, but they are not resource efficient. Incremental tree\nlearners keep adding new nodes to the tree but never remove old ones increasing\nthe memory consumption over time. Gradient-based tree learning, on the other\nhand, requires the computation of gradients over the entire tree which is\ncostly for even moderately sized trees. In this paper, we propose a novel\nmemory-efficient online classification ensemble called shrub ensembles for\nresource-constraint systems. Our algorithm trains small to medium-sized\ndecision trees on small windows and uses stochastic proximal gradient descent\nto learn the ensemble weights of these `shrubs'. We provide a theoretical\nanalysis of our algorithm and include an extensive discussion on the behavior\nof our approach in the online setting. In a series of 2~959 experiments on 12\ndifferent datasets, we compare our method against 8 state-of-the-art methods.\nOur Shrub Ensembles retain an excellent performance even when only little\nmemory is available. We show that SE offers a better accuracy-memory trade-off\nin 7 of 12 cases, while having a statistically significant better performance\nthan most other methods. Our implementation is available under\nhttps://github.com/sbuschjaeger/se-online .\n", " We provide explicit representations of Green's functions for general linear\nfractional differential operators with {\\it variable coefficients} and\nRiemann-Liouvilles derivatives. We assume that all their coefficients are\ncontinuous in $[0, \\infty)$. Using the explicit representations for Green's\nfunction, we obtain explicit representations for solution of inhomogeneous\nfractional differential equation with variable coefficients of general type.\nTherefore the method of Green's function, which was developed in previous\nresearch for solution of fractional differential equation with constant\ncoefficients, is extended to the case of fractional differential equations with\n{\\it variable coefficients}.\n", " We show that edge states similar to those known for topological insulators\nexist in two-dimensional electron system with one-band spectrum in the presence\nof heterogeneous spin-orbit interaction (SOI). These states appear at\nboundaries between regions with the SOIs of different kind or between the\nregions with the SOI and without it. Depending on the system parameters they\ncan appear in an energy range lying both in the forbidden and conduction bands\nof bulk states. The edge states have chiral spin texture and carry a spin\ncurrent under the equilibrium. We study also the size quantization of the edge\nstates in a strip structure with two boundaries to find an unusual dependence\nof the quantization energy on the strip width.\n", " The Grassmann manifold G_{n,p}(L) is the set of all p-dimensional planes\n(through the origin) in the n-dimensional Euclidean space L^{n}, where L is\neither R or C. This paper considers an unequal dimensional quantization in\nwhich a source in G_{n,p}(L) is quantized through a code in G_{n,q}(L), where p\nand q are not necessarily the same. It is different from most works in\nliterature where p\\equiv q. The analysis for unequal dimensional quantization\nis based on the volume of a metric ball in G_{n,p}(L) whose center is in\nG_{n,q}(L). Our chief result is a closed-form formula for the volume of a\nmetric ball when the radius is sufficiently small. This volume formula holds\nfor Grassmann manifolds with arbitrary n, p, q and L, while previous results\npertained only to some special cases. Based on this volume formula, several\nbounds are derived for the rate distortion tradeoff assuming the quantization\nrate is sufficiently high. The lower and upper bounds on the distortion rate\nfunction are asymptotically identical, and so precisely quantify the asymptotic\nrate distortion tradeoff. We also show that random codes are asymptotically\noptimal in the sense that they achieve the minimum achievable distortion with\nprobability one as n and the code rate approach infinity linearly. Finally, we\ndiscuss some applications of the derived results to communication theory. A\ngeometric interpretation in the Grassmann manifold is developed for capacity\ncalculation of additive white Gaussian noise channel. Further, the derived\ndistortion rate function is beneficial to characterizing the effect of\nbeamforming matrix selection in multi-antenna communications.\n", " Neural Architecture Search (NAS) has fostered the automatic discovery of\nstate-of-the-art neural architectures. Despite the progress achieved with NAS,\nso far there is little attention to theoretical guarantees on NAS. In this\nwork, we study the generalization properties of NAS under a unifying framework\nenabling (deep) layer skip connection search and activation function search. To\nthis end, we derive the lower (and upper) bounds of the minimum eigenvalue of\nthe Neural Tangent Kernel (NTK) under the (in)finite-width regime using a\ncertain search space including mixed activation functions, fully connected, and\nresidual neural networks. We use the minimum eigenvalue to establish\ngeneralization error bounds of NAS in the stochastic gradient descent training.\nImportantly, we theoretically and experimentally show how the derived results\ncan guide NAS to select the top-performing architectures, even in the case\nwithout training, leading to a train-free algorithm based on our theory.\nAccordingly, our numerical validation shed light on the design of\ncomputationally efficient methods for NAS. Our analysis is non-trivial due to\nthe coupling of various architectures and activation functions under the\nunifying framework and has its own interest in providing the lower bound of the\nminimum eigenvalue of NTK in deep learning theory.\n", " Neural network training is usually accomplished by solving a non-convex\noptimization problem using stochastic gradient descent. Although one optimizes\nover the networks parameters, the main loss function generally only depends on\nthe realization of the neural network, i.e. the function it computes. Studying\nthe optimization problem over the space of realizations opens up new ways to\nunderstand neural network training. In particular, usual loss functions like\nmean squared error and categorical cross entropy are convex on spaces of neural\nnetwork realizations, which themselves are non-convex. Approximation\ncapabilities of neural networks can be used to deal with the latter\nnon-convexity, which allows us to establish that for sufficiently large\nnetworks local minima of a regularized optimization problem on the realization\nspace are almost optimal. Note, however, that each realization has many\ndifferent, possibly degenerate, parametrizations. In particular, a local\nminimum in the parametrization space needs not correspond to a local minimum in\nthe realization space. To establish such a connection, inverse stability of the\nrealization map is required, meaning that proximity of realizations must imply\nproximity of corresponding parametrizations. We present pathologies which\nprevent inverse stability in general, and, for shallow networks, proceed to\nestablish a restricted space of parametrizations on which we have inverse\nstability w.r.t. to a Sobolev norm. Furthermore, we show that by optimizing\nover such restricted sets, it is still possible to learn any function which can\nbe learned by optimization over unrestricted sets.\n", " This paper is the first in a forthcoming series of works where the authors\nstudy the global asymptotic behavior of the radial solutions of the 2D periodic\nToda equation of type $A_n$. The principal issue is the connection formulae\nbetween the asymptotic parameters describing the behavior of the general\nsolution at zero and infinity. To reach this goal we are using a fusion of the\nPDE analysis and the Riemann-Hilbert nonlinear steepest descent method of Deift\nand Zhou which is applicable to 2D Toda in view of its Lax integrability. A\nprincipal technical challenge is the extension of the nonlinear steepest\ndescent analysis to Riemann-Hilbert problems of matrix rank greater than $2$.\nIn this paper, we meet this challenge for the case $n=2$ (the rank $3$ case)\nand it already captures the principal features of the general $n$ case.\n", " We investigate gravity as a gauge theory in the language of fiber bundles\nwith tools from algebraic geometry. Compelled by the construction of the\nEilenberg-MacLane classifying space via Fox derivations in an integral group\nring, the origin of locally curved spacetime as a an abelian group extension is\nsought algebraically by means of a suitable derivation rule that conforms with\nthe algebraic content of Einstein's equation. Accordingly, a peculiar flavor of\nbackground independence is advanced, curved spacetime should be thought of as\nan abelian proxy for the Lorentz group, causal and spin throughout. Thus,\ndiffeomorphism invariance amounts to \"choosing\" an appropriate\nalgebro-geometric--differential tool as to make algebraic sense of Einstein's\nequation or of a sensible evolution of such an equation. In this regard, it is\nsuggested that a left-$\\alpha$ derivation rule may facilitate the construction.\nOn the other hand, we are led to interpret \"spacetime curvature\" in the broad\nsense of the exterior covariant derivative of a Cartan connection which\nincludes both the torsion of the vierbein as well as the standard curvature of\nthe spin connection. Making manifest use of the index structure in the standard\nHilbert action--in particular, the distinction between internal and external\nindices--inclusion of the torsion 2-form is found to be inconsistent with\ncontraction of the spacetime curvature as a scalar density. Aiming to erect a\ngeometrically justified gravitational action that incorporates torsion, we lay\nout several fundamental geometric, differential and algebraic building blocks\nwhile highlighting descent relations among the different pieces. The groundwork\nis thus laid out for the search of the sought after action functional.\n", " We present a minimal model for simulating dynamics of assorted lipid\nassemblies in a computationally efficient manner. Our model is particle-based\nand consists of coarse-grained beads put together on a modular platform to give\ngeneric molecular lipids with tunable properties. The interaction between\ncoarse-grained beads is governed by soft, short-ranged potentials that account\nfor the renormalized hydrophobic effect in implicit solvent. The model\nfaithfully forms micelles, tubular micelles, or bilayers (periodically\ninfinite, bicelles, or vesicles) depending on the packing ratio of the lipid\nmolecules. Importantly, for the self-assembled bilayer membranes it is\nstraightforward to realize gel and fluid phases, the latter of which is most\noften of primary interest. We show that the emerging physics over a wide range\nof scales and resolutions is (with some restrictions) universal. The model,\nwhen compared to other popular lipid models, demonstrates improvements in the\nform of increased numerical stability and boosted dynamics. Possible strategies\nfor customizing the models (e.g., adding chemical specificity) are briefly\ndiscussed. An implementation is available for the LAMMPS molecular dynamics\nsimulator [Plimpton. J. Comp. Phys. 117, 1-19. (1995)] including illustrative\ninput examples from the simulations we present.\n", " We study the quantization of the motion of a charged particle without spin\ninside a flat box under a static electromagnetic field. Contrary to Landau's\nsolution with constant magnetic field transverse to the box, we found a non\nseparable variables solution for the wave function, and this fact remains when\nstatic electric field is added. However, the Landau's Levels appear in all\ncases.\n", " Curvature principal directions on geometric surfaces are a ubiquitous concept\nof Geometry Processing techniques. However they only account for order 2\ndifferential quantities, oblivious of higher order differential behaviors. In\nthis paper, we extend the concept of principal directions to higher orders for\nsurfaces in R^3 by considering symmetric differential tensors. We further show\nhow they can be explicitly approximated on point set surfaces and that they\nconvey valuable geometric information, that can help the analysis of 3D\nsurfaces.\n", " Today, many users deploy their microservice-based applications with various\ninterconnections on a cluster of Cloud machines, subject to stochastic changes\ndue to dynamic user requirements. To address this problem, we compare three\nmachine learning (ML) models for predicting the microservice call rates based\non the microservice times and aiming at estimating the scalability\nrequirements. We apply the linear regression (LR), multilayer perception (MLP),\nand gradient boosting regression (GBR) models on the Alibaba microservice\ntraces. The prediction results reveal that the LR model reaches a lower\ntraining time than the GBR and MLP models. However, the GBR reduces the mean\nabsolute error and the mean absolute percentage error compared to LR and MLP\nmodels. Moreover, the prediction results show that the required number of\nreplicas for each microservice by the gradient boosting model is close to the\nactual test data without any prediction.\n", " Using geometric methods for linearizing systems of second order cubically\nsemi-linear ordinary differential equations, we extend to the third order by\ndifferentiating the second order equation. This yields criteria for\nlinearizability of a class of third order semi-linear ordinary differential\nequations, which is distinct from the classes available in the literature. Some\nexamples are given and discussed.\n", " We outline a connection between scalar quark confinement, a\nphenomenologically successful concept heretofore lacking fundamental\njustification, and QCD. Although scalar confinement does not follow from QCD,\nthere is an interesting and close relationship between them. We develop a\nsimple model intermediate between scalar confinement and the QCD string for\nillustrative purposes. Finally, we find the bound state masses of scalar,\ntime-component vector, and string confinement analytically through\nsemi-classical quantization.\n", " A result of Foata and Schutzenberger states that two statistics on\npermutations, the number of inversions and the inverse major index, have the\nsame distribution on a descent class. We give a multivariate generalization of\nthis property: the sorted vectors of the Lehmer code, of the inverse majcode,\nand of a new code (the inverse saillance code), have the same distribution on a\ndescent class, and their common multivariate generating function is a flagged\nribbon Schur function.\n", " In this article, we recover singularly-perturbed linear differential systems\nfrom their turning points and reduce the rank of the singularity in the\nparameter to its minimal integer value. Our treatment is Moser-based; that is\nto say it is based on the reduction criterion introduced for linear singular\ndifferential systems by Moser. Such algorithms have proved their utility in the\nsymbolic resolution of the systems of linear functional equations, giving rise\nto the package ISOLDE, as well as in the perturbed algebraic eigenvalue\nproblem. Our algorithm, implemented in the computer algebra system Maple, paves\nthe way for efficient symbolic resolution of singularly-perturbed linear\ndifferential systems as well as further applications of Moser-based reduction\nover bivariate (differential) fields.\n", " We propose a linear time and constant space algorithm for computing Euclidean\nprojections onto sets on which a normalized sparseness measure attains a\nconstant value. These non-convex target sets can be characterized as\nintersections of a simplex and a hypersphere. Some previous methods required\nthe vector to be projected to be sorted, resulting in at least quasilinear time\ncomplexity and linear space complexity. We improve on this by adaptation of a\nlinear time algorithm for projecting onto simplexes. In conclusion, we propose\nan efficient algorithm for computing the product of the gradient of the\nprojection with an arbitrary vector.\n", " D-Adaptation is an approach to automatically setting the learning rate which\nasymptotically achieves the optimal rate of convergence for minimizing convex\nLipschitz functions, with no back-tracking or line searches, and no additional\nfunction value or gradient evaluations per step. Our approach is the first\nhyper-parameter free method for this class without additional multiplicative\nlog factors in the convergence rate. We present extensive experiments for SGD\nand Adam variants of our method, where the method automatically matches\nhand-tuned learning rates across more than a dozen diverse machine learning\nproblems, including large-scale vision and language problems.\n An open-source implementation is available.\n", " We introduce an improved semiclassical dynamics approach to quantum\nvibrational spectroscopy. In this method, a harmonic-based phase space sampling\nis preliminarily driven toward non-harmonic quantization by slowly switching on\nthe actual potential. The new coordinates and momenta serve as initial\nconditions for the semiclassical dynamics calculation, leading to a substantial\ndecrease in the number of chaotic trajectories to deal with. Applications are\npresented for model and molecular systems of increasing dimensionality\ncharacterized by moderate or high chaoticity. They include a bidimensional\nHenon-Heiles potential, water, formaldehyde, and methane.The method improves\naccuracy and precision of semiclassical results and it can be easily interfaced\nwith all pre-existing semiclassical theories.\n", " The differential transformation method (DTM) enables the easy construction of\na power-series solution to a nonlinear differential equation. The\nexponentiation operation has not been specifically addressed in the DTM\nliterature, and constructing it iteratively is suboptimal. The recurrence for\nexponentiating a power series by J.C.P. Miller provides a concise\nimplementation of exponentiation by a positive integer for DTM. An\nequally-concise implementation of the exponential function is also provided.\n" ]

[ " In this work we include, for the Carnot cycle, irreversibilities of linear\nfinite rate of heat transferences between the heat engine and its reservoirs,\nheat leak between the reservoirs and internal dissipations of the working\nfluid. A first optimization of the power output, the efficiency and ecological\nfunction of an irreversible Carnot cycle, with respect to: internal temperature\nratio, time ratio for the heat exchange and the allocation ratio of the heat\nexchangers; is performed. For the second and third optimizations, the optimum\nvalues for the time ratio and internal temperature ratio are substituted into\nthe equation of power and, then, the optimizations with respect to the cost and\neffectiveness ratio of the heat exchangers are performed. Finally, a criterion\nof partial optimization for the class of irreversible Carnot engines is herein\npresented.\n", " In this paper we analyze the performance of 2*4 MIMO-MC-CDMA system in MATLAB\nwhich highly reduces BER. In this paper we combine MIMO and MC-CDMA system to\nreduce bit error rate in which MC-CDMA is multi user and multiple access\nschemes which is used to increase the data rate of the system. MC-CDMA system\nis a single wideband frequency selective carrier which converts frequency\nselective to parallel narrowband flat fading multiple sub-carriers to enhance\nthe performance of system. Now MC-CDMA system further improved by grouping with\n2*4 MIMO system which uses ZF (Zero Forcing) decoder at the receiver to\ndecrease BER with half rate convolutionally encoded Alamouti STBC block code is\nused as transmit diversity of MIMO through multiple transmit antenna.\nImportance of using MIMO-MC-CDMA using convolution code is firstly to reduce\nthe complexity of system secondary to reduce BER and lastly to increase gain.\nIn this paper we examine system performance in diverse modulation techniques\nlike, 8-PSK, 16-QAM, QPSK, 32-QAM, 8-QAM and 64-QAM in Rayleigh fading channel\nusing MATLAB.\n", " Within a fully microscopic setting, we derive a variational principle for the\nnon-equilibrium steady states of chemical reaction networks, valid for\ntime-scales over which chemical potentials can be taken to be slowly varying:\nat stationarity the system minimizes a global function of the reaction fluxes\nwith the form of a Hopfield Hamiltonian with Hebbian couplings, that is\nexplicitly seen to correspond to the rate of decay of entropy production over\ntime. Guided by this analogy, we show that reaction networks can be formally\nre-cast as systems of interacting reactions that optimize the use of the\navailable compounds by competing for substrates, akin to agents competing for a\nlimited resource in an optimal allocation problem. As an illustration, we\nanalyze the scenario that emerges in two simple cases: that of toy (random)\nreaction networks and that of a metabolic network model of the human red blood\ncell.\n", " We propose a framework, called neural-progressive hedging (NP), that\nleverages stochastic programming during the online phase of executing a\nreinforcement learning (RL) policy. The goal is to ensure feasibility with\nrespect to constraints and risk-based objectives such as conditional\nvalue-at-risk (CVaR) during the execution of the policy, using probabilistic\nmodels of the state transitions to guide policy adjustments. The framework is\nparticularly amenable to the class of sequential resource allocation problems\nsince feasibility with respect to typical resource constraints cannot be\nenforced in a scalable manner. The NP framework provides an alternative that\nadds modest overhead during the online phase. Experimental results demonstrate\nthe efficacy of the NP framework on two continuous real-world tasks: (i) the\nportfolio optimization problem with liquidity constraints for financial\nplanning, characterized by non-stationary state distributions; and (ii) the\ndynamic repositioning problem in bike sharing systems, that embodies the class\nof supply-demand matching problems. We show that the NP framework produces\npolicies that are better than deep RL and other baseline approaches, adapting\nto non-stationarity, whilst satisfying structural constraints and accommodating\nrisk measures in the resulting policies. Additional benefits of the NP\nframework are ease of implementation and better explainability of the policies.\n", " When the users in a MIMO broadcast channel experience different spatial\ntransmit correlation matrices, a class of gains is produced that is denoted\ntransmit correlation diversity. This idea was conceived for channels in which\ntransmit correlation matrices have mutually exclusive eigenspaces, allowing\nnon-interfering training and transmission. This paper broadens the scope of\ntransmit correlation diversity to the case of partially and fully overlapping\neigenspaces and introduces techniques to harvest these generalized gains. For\nthe two-user MIMO broadcast channel, we derive achievable degrees of freedom\n(DoF) and achievable rate regions with/without channel state information at the\nreceiver (CSIR). When CSIR is available, the proposed achievable DoF region is\ntight in some configurations of the number of receive antennas and the channel\ncorrelation ranks. We then extend the DoF results to the $K$-user case by\nanalyzing the interference graph that characterizes the overlapping structure\nof the eigenspaces. Our achievability results employ a combination of product\nsuperposition in the common part of the eigenspaces, and pre-beamforming (rate\nsplitting) to create multiple data streams in non-overlapping parts of the\neigenspaces. Massive MIMO is a natural example in which spatially correlated\nlink gains are likely to occur. We study the achievable downlink sum rate for a\nfrequency-division duplex massive MIMO system under transmit correlation\ndiversity.\n", " With the rapid advancement of 5G networks, billions of smart Internet of\nThings (IoT) devices along with an enormous amount of data are generated at the\nnetwork edge. While still at an early age, it is expected that the evolving 6G\nnetwork will adopt advanced artificial intelligence (AI) technologies to\ncollect, transmit, and learn this valuable data for innovative applications and\nintelligent services. However, traditional machine learning (ML) approaches\nrequire centralizing the training data in the data center or cloud, raising\nserious user-privacy concerns. Federated learning, as an emerging distributed\nAI paradigm with privacy-preserving nature, is anticipated to be a key enabler\nfor achieving ubiquitous AI in 6G networks. However, there are several system\nand statistical heterogeneity challenges for effective and efficient FL\nimplementation in 6G networks. In this article, we investigate the optimization\napproaches that can effectively address the challenging heterogeneity issues\nfrom three aspects: incentive mechanism design, network resource management,\nand personalized model optimization. We also present some open problems and\npromising directions for future research.\n", " Several adaptation techniques have been investigated to optimize fuzzy\ninference systems. Neural network learning algorithms have been used to\ndetermine the parameters of fuzzy inference system. Such models are often\ncalled as integrated neuro-fuzzy models. In an integrated neuro-fuzzy model\nthere is no guarantee that the neural network learning algorithm converges and\nthe tuning of fuzzy inference system will be successful. Success of\nevolutionary search procedures for optimization of fuzzy inference system is\nwell proven and established in many application areas. In this paper, we will\nexplore how the optimization of fuzzy inference systems could be further\nimproved using a meta-heuristic approach combining neural network learning and\nevolutionary computation. The proposed technique could be considered as a\nmethodology to integrate neural networks, fuzzy inference systems and\nevolutionary search procedures. We present the theoretical frameworks and some\nexperimental results to demonstrate the efficiency of the proposed technique.\n", " This paper deals with temporal enzyme distribution in the activation of\nbiochemical pathways. Pathway activation arises when production of a certain\nbiomolecule is required due to changing environmental conditions. Under the\npremise that biological systems have been optimized through evolutionary\nprocesses, a biologically meaningful optimal control problem is posed. In this\nsetup, the enzyme concentrations are assumed to be time dependent and\nconstrained by a limited overall enzyme production capacity, while the\noptimization criterion accounts for both time and resource usage.\n Using geometric arguments we establish the bang-bang nature of the solution\nand reveal that each reaction must be sequentially activated in the same order\nas they appear in the pathway. The results hold for a broad range of enzyme\ndynamics which includes, but is not limited to, Mass Action, Michaelis-Menten\nand Hill Equation kinetics.\n", " The ability to shop independently, especially in grocery stores, is important\nfor maintaining a high quality of life. This can be particularly challenging\nfor people with visual impairments (PVI). Stores carry thousands of products,\nwith approximately 30,000 new products introduced each year in the US market\nalone, presenting a challenge even for modern computer vision solutions.\nThrough this work, we present a proof-of-concept socially assistive robotic\nsystem we call ShelfHelp, and propose novel technical solutions for enhancing\ninstrumented canes traditionally meant for navigation tasks with additional\ncapability within the domain of shopping. ShelfHelp includes a novel visual\nproduct locator algorithm designed for use in grocery stores and a novel\nplanner that autonomously issues verbal manipulation guidance commands to guide\nthe user during product retrieval. Through a human subjects study, we show the\nsystem's success in locating and providing effective manipulation guidance to\nretrieve desired products with novice users. We compare two autonomous verbal\nguidance modes achieving comparable performance to a human assistance baseline\nand present encouraging findings that validate our system's efficiency and\neffectiveness and through positive subjective metrics including competence,\nintelligence, and ease of use.\n", " Prediction+optimization is a common real-world paradigm where we have to\npredict problem parameters before solving the optimization problem. However,\nthe criteria by which the prediction model is trained are often inconsistent\nwith the goal of the downstream optimization problem. Recently,\ndecision-focused prediction approaches, such as SPO+ and direct optimization,\nhave been proposed to fill this gap. However, they cannot directly handle the\nsoft constraints with the $max$ operator required in many real-world\nobjectives. This paper proposes a novel analytically differentiable surrogate\nobjective framework for real-world linear and semi-definite negative quadratic\nprogramming problems with soft linear and non-negative hard constraints. This\nframework gives the theoretical bounds on constraints' multipliers, and derives\nthe closed-form solution with respect to predictive parameters and thus\ngradients for any variable in the problem. We evaluate our method in three\napplications extended with soft constraints: synthetic linear programming,\nportfolio optimization, and resource provisioning, demonstrating that our\nmethod outperforms traditional two-staged methods and other decision-focused\napproaches.\n", " Deep-learning-based intelligent services have become prevalent in\ncyber-physical applications including smart cities and health-care. Deploying\ndeep-learning-based intelligence near the end-user enhances privacy protection,\nresponsiveness, and reliability. Resource-constrained end-devices must be\ncarefully managed in order to meet the latency and energy requirements of\ncomputationally-intensive deep learning services. Collaborative end-edge-cloud\ncomputing for deep learning provides a range of performance and efficiency that\ncan address application requirements through computation offloading. The\ndecision to offload computation is a communication-computation co-optimization\nproblem that varies with both system parameters (e.g., network condition) and\nworkload characteristics (e.g., inputs). On the other hand, deep learning model\noptimization provides another source of tradeoff between latency and model\naccuracy. An end-to-end decision-making solution that considers such\ncomputation-communication problem is required to synergistically find the\noptimal offloading policy and model for deep learning services. To this end, we\npropose a reinforcement-learning-based computation offloading solution that\nlearns optimal offloading policy considering deep learning model selection\ntechniques to minimize response time while providing sufficient accuracy. We\ndemonstrate the effectiveness of our solution for edge devices in an\nend-edge-cloud system and evaluate with a real-setup implementation using\nmultiple AWS and ARM core configurations. Our solution provides 35% speedup in\nthe average response time compared to the state-of-the-art with less than 0.9%\naccuracy reduction, demonstrating the promise of our online learning framework\nfor orchestrating DL inference in end-edge-cloud systems.\n", " Stochastic gradient descent (SGD) is a powerful method for large-scale\noptimization problems in the area of machine learning, especially for a\nfinite-sum formulation with numerous variables. In recent years, mini-batch SGD\ngains great success and has become a standard technique for training deep\nneural networks fed with big amount of data. Inspired by its success in deep\nlearning, we apply the idea of SGD with batch selection of samples to a classic\noptimization problem in decision version. Given $n$ unit circles, the equal\ncircle packing problem (ECPP) asks whether there exist a feasible packing that\ncould put all the circles inside a circular container without overlapping.\nSpecifically, we propose a stochastic item descent method (SIDM) for ECPP in\nlarge scale, which randomly divides the unit circles into batches and runs\nBroyden-Fletcher-Goldfarb-Shanno (BFGS) algorithm on the corresponding batch\nfunction iteratively to speedup the calculation. We also increase the batch\nsize during the batch iterations to gain higher quality solution. Comparing to\nthe current best packing algorithms, SIDM greatly speeds up the calculation of\noptimization process and guarantees the solution quality for large scale\ninstances with up to 1500 circle items, while the baseline algorithms usually\nhandle about 300 circle items. The results indicate the highly efficiency of\nSIDM for this classic optimization problem in large scale, and show potential\nfor other large scale classic optimization problems in which gradient descent\nis used for optimization.\n", " We study a fundamental NP-hard motion coordination problem for\nmulti-robot/multi-agent systems: We are given a graph $G$ and set of agents,\nwhere each agent has a given directed path in $G$. Each agent is initially\nlocated on the first vertex of its path. At each time step an agent can move to\nthe next vertex on its path, provided that the vertex is not occupied by\nanother agent. The goal is to find a sequence of such moves along the given\npaths so that each reaches its target, or to report that no such sequence\nexists. The problem models guidepath-based transport systems, which is a\npertinent abstraction for traffic in a variety of contemporary applications,\nranging from train networks or Automated Guided Vehicles (AGVs) in factories,\nthrough computer game animations, to qubit transport in quantum computing. It\nalso arises as a sub-problem in the more general multi-robot motion-planning\nproblem.\n We provide a fine-grained tractability analysis of the problem by considering\nnew assumptions and identifying minimal values of key parameters for which the\nproblem remains NP-hard. Our analysis identifies a critical parameter called\nvertex multiplicity (VM), defined as the maximum number of paths passing\nthrough the same vertex. We show that a prevalent variant of the problem, which\nis equivalent to Sequential Resource Allocation (concerning deadlock prevention\nfor concurrent processes), is NP-hard even when VM is 3. On the positive side,\nfor VM $\\le$ 2 we give an efficient algorithm that iteratively resolves cycles\nof blocking relations among agents. We also present a variant that is NP-hard\nwhen the VM is 2 even when $G$ is a 2D grid and each path lies in a single grid\nrow or column. By studying highly distilled yet NP-hard variants, we deepen the\nunderstanding of what makes the problem intractable and thereby guide the\nsearch for efficient solutions under practical assumptions.\n", " In this study, a hybrid method based on an Adaptive Neuro-Fuzzy Inference\nSystem (ANFIS) and Particle Swarm Optimization (PSO) for diagnosing Liver\ndisorders (ANFIS-PSO) is introduced. This smart diagnosis method deals with a\ncombination of making an inference system and optimization process which tries\nto tune the hyper-parameters of ANFIS based on the data-set. The Liver diseases\ncharacteristics are taken from the UCI Repository of Machine Learning\nDatabases. The number of these characteristic attributes are 7, and the sample\nnumber is 354. The right diagnosis performance of the ANFIS-PSO intelligent\nmedical system for liver disease is evaluated by using classification accuracy,\nsensitivity and specificity analysis, respectively. According to the\nexperimental results, the performance of ANFIS-PSO can be more considerable\nthan traditional FIS and ANFIS without optimization phase.\n", " A generic and intuitive model for coherent energy transport in multiple\nminima systems coupled to a quantum mechanical bath is shown. Using a simple\nspin-boson system, we illustrate how a generic donor-acceptor system can be\nbrought into resonance using a narrow band of vibrational modes, such that the\ntransfer efficiency of an electron-hole pair (exciton) is made arbitrarily\nhigh. Coherent transport phenomena in nature are of renewed interest since the\ndiscovery that a photon captured by the light-harvesting complex (LHC) in\nphotosynthetic organisms can be conveyed to a chemical reaction centre with\nnear-perfect efficiency. Classical explanations of the transfer use stochastic\ndiffusion to model the hopping motion of a photo-excited exciton. This accounts\ninadequately for the speed and efficiency of the energy transfer measured in a\nseries of recent landmark experiments. Taking a quantum mechanical perspective\ncan help capture the salient features of the efficient part of that transfer.\nTo show the versatility of the model, we extend it to a multiple minima system\ncomprising seven-sites, reminiscent of the widely studied Fenna-Matthews-Olson\n(FMO) light-harvesting complex. We show that an idealised transport model for\nmultiple minima coupled to a narrow-band phonon can transport energy with\narbitrarily high efficiency.\n", " Bilevel Optimization Programming is used to model complex and conflicting\ninteractions between agents, for example in Robust AI or Privacy-preserving AI.\nIntegrating bilevel mathematical programming within deep learning is thus an\nessential objective for the Machine Learning community. Previously proposed\napproaches only consider single-level programming. In this paper, we extend\nexisting single-level optimization programming approaches and thus propose\nDifferentiating through Bilevel Optimization Programming (BiGrad) for\nend-to-end learning of models that use Bilevel Programming as a layer. BiGrad\nhas wide applicability and can be used in modern machine learning frameworks.\nBiGrad is applicable to both continuous and combinatorial Bilevel optimization\nproblems. We describe a class of gradient estimators for the combinatorial case\nwhich reduces the requirements in terms of computation complexity; for the case\nof the continuous variable, the gradient computation takes advantage of the\npush-back approach (i.e. vector-jacobian product) for an efficient\nimplementation. Experiments show that the BiGrad successfully extends existing\nsingle-level approaches to Bilevel Programming.\n", " The availability of accurate localization is critical for multi-robot\nexploration strategies; noisy or inconsistent localization causes failure in\nmeeting exploration objectives. We aim to achieve high localization accuracy\nwith contemporary exploration map belief and vice versa without needing global\nlocalization information. This paper proposes a novel simultaneous exploration\nand localization (SEAL) approach, which uses Gaussian Processes (GP)-based\ninformation fusion for maximum exploration while performing communication graph\noptimization for relative localization. Both these cross-dependent objectives\nwere integrated through the Rao-Blackwellization technique. Distributed\nlinearized convex hull optimization is used to select the next-best unexplored\nregion for distributed exploration. SEAL outperformed cutting-edge methods on\nexploration and localization performance in extensive ROS-Gazebo simulations,\nillustrating the practicality of the approach in real-world applications.\n", " We demonstrate that, for a chemical reaction network (CRN) engaged in energy\ntransduction, its optimal operation from a thermodynamic efficiency standpoint\nis contingent upon its working conditions. Analogously to the bicycle gear\nsystem, CRNs have at their disposal several transducing mechanisms\ncharacterized by different yields. We highlight the critical role of the CRN's\nelementary flux modes in determining this \"gearing\" and their impact on\nmaximizing energy transduction efficiency. Furthermore, we introduce an\nenzymatically regulated CRN, engineered to autonomously adjust its \"gear\",\nthereby optimizing its efficiency under different external conditions.\n", " One of the first widespread uses of multi-user multiple-input multiple-output\n(MU-MIMO) is in 5G networks, where each base station has an advanced antenna\nsystem (AAS) that is connected to the baseband unit (BBU) with a\ncapacity-constrained fronthaul. In the AAS configuration, multiple passive\nantenna elements and radio units are integrated into a single box. This paper\nconsiders precoded downlink transmission over a single-cell MU-MIMO system. We\nstudy optimized linear precoding for AAS with a limited-capacity fronthaul,\nwhich requires the precoding matrix to be quantized. We propose a new precoding\ndesign that is aware of the fronthaul quantization and minimizes the\nmean-squared error at the receiver side. We compute the precoding matrix using\na sphere decoding (SD) approach. We also propose a heuristic low-complexity\napproach to quantized precoding. This heuristic is computationally efficient\nenough for massive MIMO systems. The numerical results show that our proposed\nprecoding significantly outperforms quantization-unaware precoding and other\nprevious approaches in terms of the sum rate. The performance loss for our\nheuristic method compared to quantization-aware precoding is insignificant\nconsidering the complexity reduction, which makes the heuristic method feasible\nfor real-time applications. We consider both perfect and imperfect channel\nstate information.\n", " We explore the interplay between the protein-protein interactions network and\nthe expression of the interacting proteins. It is shown that interacting\nproteins are expressed in significantly more similar cellular concentrations.\nThis is largely due to interacting pairs which are part of protein complexes.\nWe solve a generic model of complex formation and show explicitly that\ncomplexes form most efficiently when their members have roughly the same\nconcentrations. Therefore, the observed similarity in interacting protein\nconcentrations could be attributed to optimization for efficiency of complex\nformation.\n", " In this paper, we tested several sparse optimization algorithms based on the\npublic dataset of the DREAM5 Gene Regulatory Network Inference Challenge. And\nwe find that introducing 20% of the regulatory network as a priori known data\ncan provide a basis for parameter selection of inference algorithms, thus\nimproving prediction efficiency and accuracy. In addition to testing common\nsparse optimization methods, we also developed voting algorithms by bagging\nthem. Experiments on the DREAM5 dataset show that the sparse optimization-based\ninference of the moderation relation works well, achieving better results than\nthe official DREAM5 results on three datasets. However, the performance of\ntraditional independent algorithms varies greatly in the face of different\ndatasets, while our voting algorithm achieves the best results on three of the\nfour datasets.\n", " Large-scale multi-modal deep learning models have revolutionized domains such\nas healthcare, highlighting the importance of computational power. However, in\nresource-constrained regions like Low and Middle-Income Countries (LMICs),\nlimited access to GPUs and data poses significant challenges, often leaving\nCPUs as the sole resource. To address this, we advocate for leveraging vector\nembeddings to enable flexible and efficient computational methodologies,\ndemocratizing multimodal deep learning across diverse contexts.\n Our paper investigates the efficiency and effectiveness of using vector\nembeddings from single-modal foundation models and multi-modal Vision-Language\nModels (VLMs) for multimodal deep learning in low-resource environments,\nparticularly in healthcare. Additionally, we propose a simple yet effective\ninference-time method to enhance performance by aligning image-text embeddings.\nComparing these approaches with traditional methods, we assess their impact on\ncomputational efficiency and model performance using metrics like accuracy,\nF1-score, inference time, training time, and memory usage across three medical\nmodalities: BRSET (ophthalmology), HAM10000 (dermatology), and SatelliteBench\n(public health).\n Our findings show that embeddings reduce computational demands without\ncompromising model performance. Furthermore, our alignment method improves\nperformance in medical tasks. This research promotes sustainable AI practices\nby optimizing resources in constrained environments, highlighting the potential\nof embedding-based approaches for efficient multimodal learning. Vector\nembeddings democratize multimodal deep learning in LMICs, particularly in\nhealthcare, enhancing AI adaptability in varied use cases.\n", " We develop an efficient algorithm, coordinate descent FCI (CDFCI), for the\nelectronic structure ground state calculation in the configuration interaction\nframework. CDFCI solves an unconstrained non-convex optimization problem, which\nis a reformulation of the full configuration interaction eigenvalue problem,\nvia an adaptive coordinate descent method with a deterministic compression\nstrategy. CDFCI captures and updates appreciative determinants with different\nfrequencies proportional to their importance. We show that CDFCI produces\naccurate variational energy for both static and dynamic correlation by\nbenchmarking the binding curve of nitrogen dimer in the cc-pVDZ basis with\n$10^{-3}$ mHa accuracy. We also demonstrate the efficiency and accuracy of\nCDFCI for strongly correlated chromium dimer in the Ahlrichs VDZ basis and\nproduces state-of-the-art variational energy.\n", " Living organisms continuously harness energy to perform complex functions for\ntheir adaptation and survival while part of that energy is dissipated in the\nform of heat or chemical waste. Determining the energetic cost and the\nefficiency of specific cellular processes remains a largely open problem. Here,\nwe analyze the efficiency of mitochondrial adenosine triphosphate (ATP)\nproduction through the tricarboxylic acid (TCA) cycle and oxidative\nphosphorylation that generates most of the cellular chemical energy in\neukaryotes. The regulation of this pathway by calcium signaling represents a\nwell-characterized example of a regulatory cross-talk that can affect the\nenergetic output of a metabolic pathway, but its concrete energetic impact\nremains elusive. On the one hand, calcium enhances ATP production by activating\nkey enzymes of the TCA cycle, but on the other hand calcium homeostasis depends\non ATP availability. To evaluate how calcium signaling impacts the efficiency\nof mitochondrial metabolism, we propose a detailed kinetic model describing the\ncalcium-mitochondria cross-talk and we analyze it using a nonequilibrium\nthermodynamic approach: after identifying the effective reactions driving\nmitochondrial metabolism out of equilibrium, we quantify the thermodynamic\nefficiency of the metabolic machinery for different physiological conditions.\nWe find that calcium oscillations increase the efficiency with a maximum close\nto substrate-limited conditions, suggesting a compensatory effect of calcium\nsignaling on the energetics of mitochondrial metabolism.\n", " This paper proposes a unique active relative localization mechanism for\nmulti-agent Simultaneous Localization and Mapping(SLAM),in which a agent to be\nobserved are considered as a task, which is performed by others assisting that\nagent by relative observation. A task allocation algorithm based on deep\nreinforcement learning are proposed for this mechanism. Each agent can choose\nwhether to localize other agents or to continue independent SLAM on it own\ninitiative. By this way, the process of each agent SLAM will be interacted by\nthe collaboration. Firstly, based on the characteristics of ORBSLAM, a unique\nobservation function which models the whole MAS is obtained. Secondly, a novel\ntype of Deep Q network(DQN) called MAS-DQN is deployed to learn correspondence\nbetween Q Value and state-action pair,abstract representation of agents in MAS\nare learned in the process of collaboration among agents. Finally, each agent\nmust act with a certain degree of freedom according to MAS-DQN. The simulation\nresults of comparative experiments prove that this mechanism improves the\nefficiency of cooperation in the process of multi-agent SLAM.\n", " Fault tolerance is increasingly important for unmanned autonomous vehicles.\nFor example, in a multi robot system the agents need the ability to effectively\ndetect and tolerate internal failures in order to continue performing their\ntasks without the need for immediate human intervention. The system must react\nto unplanned events in order to optimize the task allocation between the\nrobots. In a broad sense, the resilience of a system can be defined as the\nability to maintain or recover a stable state when subject to disturbance and\nit is related to the concept of robustness in industrial systems. In this\npaper, we study the resilience in a synchronized multi-robot system stated as\nfollows:Consider a team of $n$ (ground or aerial) robots each moving along\npredetermined periodic closed trajectories. Each of the agents needs to\ncommunicate informationabout its operation to other agents, but the\ncommunication links have a limited range. Hence, when two agents are within\ncommunication range, a communication link is established, and information is\nexchanged. Thus, two neighbors are synchronized if they visit the communication\nlink at the same time and a multi-robot system is called synchronized if each\npair of neighbors is synchronized. If a set of robots left the system, then\nsome trajectories has no robots. In these cases, when an alive robot detects no\nneighboring robot then it pass to this neighboring trajectory to assume the\nunattended task. In this framework, a fault-tolerance measure is introduced:\nthe resilience of the system is the largest number of robots that can fail\nwhile executing the global task. Interesting combinatorial properties of the\nresilience are showed that allow to know its value for some usual scenarios.\n", " Bacterial sensor systems can be used for the detection and measurement of\nmolecular signal concentrations. The dynamics of the sensor directly depend on\nthe biological properties of the bacterial sensor cells; manipulation of these\nfeatures in the wet lab enables the engineering and optimization of the\nbacterial sensor kinetics. This necessitates the development of biologically\nmeaningful computational models for bacterial sensors comprising a variety of\ndifferent molecular mechanisms, which further facilitates a systematic and\nquantitative evaluation of optimization strategies. In this work, we dissect\nthe detection chain of bacterial sensors from a mathematical perspective from\nwhich we derive, supported by wet-lab data, a complete computational model for\na Streptococcus mutans-based bacterial sensor as a case example. We address the\nengineering of bacterial sensors by investigating the impact of altered\nbacterial cell properties on the sensor response characteristics, specifically\nsensor sensitivity and response signal intensity. This is achieved through a\nsensitivity analysis targeting both the steady-state and transient sensor\nresponse characteristics. Alongside the demonstration of suitability of our\nmethodological approach, our analysis shows that an increase of sensor\nsensitivity, through a targeted manipulation of bacterial physiology, often\ncomes at the cost of generally diminished sensor response intensity.\n", " Current research on Internet of Things (IoT) mainly focuses on how to enable\ngeneral objects to see, hear, and smell the physical world for themselves, and\nmake them connected to share the observations. In this paper, we argue that\nonly connected is not enough, beyond that, general objects should have the\ncapability to learn, think, and understand both physical and social worlds by\nthemselves. This practical need impels us to develop a new paradigm, named\nCognitive Internet of Things (CIoT), to empower the current IoT with a `brain'\nfor high-level intelligence. Specifically, we first present a comprehensive\ndefinition for CIoT, primarily inspired by the effectiveness of human\ncognition. Then, we propose an operational framework of CIoT, which mainly\ncharacterizes the interactions among five fundamental cognitive tasks:\nperception-action cycle, massive data analytics, semantic derivation and\nknowledge discovery, intelligent decision-making, and on-demand service\nprovisioning. Furthermore, we provide a systematic tutorial on key enabling\ntechniques involved in the cognitive tasks. In addition, we also discuss the\ndesign of proper performance metrics on evaluating the enabling techniques.\nLast but not least, we present the research challenges and open issues ahead.\nBuilding on the present work and potentially fruitful future studies, CIoT has\nthe capability to bridge the physical world (with objects, resources, etc.) and\nthe social world (with human demand, social behavior, etc.), and enhance smart\nresource allocation, automatic network operation, and intelligent service\nprovisioning.\n", " Federated Learning (FL) has emerged as a promising technique for edge devices\nto collaboratively learn a shared machine learning model while keeping training\ndata locally on the device, thereby removing the need to store and access the\nfull data in the cloud. However, FL is difficult to implement, test and deploy\nin practice considering heterogeneity in common edge device settings, making it\nfundamentally hard for researchers to efficiently prototype and test their\noptimization algorithms. In this work, our aim is to alleviate this problem by\nintroducing FL_PyTorch : a suite of open-source software written in python that\nbuilds on top of one the most popular research Deep Learning (DL) framework\nPyTorch. We built FL_PyTorch as a research simulator for FL to enable fast\ndevelopment, prototyping and experimenting with new and existing FL\noptimization algorithms. Our system supports abstractions that provide\nresearchers with a sufficient level of flexibility to experiment with existing\nand novel approaches to advance the state-of-the-art. Furthermore, FL_PyTorch\nis a simple to use console system, allows to run several clients simultaneously\nusing local CPUs or GPU(s), and even remote compute devices without the need\nfor any distributed implementation provided by the user. FL_PyTorch also offers\na Graphical User Interface. For new methods, researchers only provide the\ncentralized implementation of their algorithm. To showcase the possibilities\nand usefulness of our system, we experiment with several well-known\nstate-of-the-art FL algorithms and a few of the most common FL datasets.\n", " Modern vehicles, including connected vehicles and autonomous vehicles,\nnowadays involve many electronic control units connected through intra-vehicle\nnetworks to implement various functionalities and perform actions. Modern\nvehicles are also connected to external networks through vehicle-to-everything\ntechnologies, enabling their communications with other vehicles,\ninfrastructures, and smart devices. However, the improving functionality and\nconnectivity of modern vehicles also increase their vulnerabilities to\ncyber-attacks targeting both intra-vehicle and external networks due to the\nlarge attack surfaces. To secure vehicular networks, many researchers have\nfocused on developing intrusion detection systems (IDSs) that capitalize on\nmachine learning methods to detect malicious cyber-attacks. In this paper, the\nvulnerabilities of intra-vehicle and external networks are discussed, and a\nmulti-tiered hybrid IDS that incorporates a signature-based IDS and an\nanomaly-based IDS is proposed to detect both known and unknown attacks on\nvehicular networks. Experimental results illustrate that the proposed system\ncan detect various types of known attacks with 99.99% accuracy on the\nCAN-intrusion-dataset representing the intra-vehicle network data and 99.88%\naccuracy on the CICIDS2017 dataset illustrating the external vehicular network\ndata. For the zero-day attack detection, the proposed system achieves high\nF1-scores of 0.963 and 0.800 on the above two datasets, respectively. The\naverage processing time of each data packet on a vehicle-level machine is less\nthan 0.6 ms, which shows the feasibility of implementing the proposed system in\nreal-time vehicle systems. This emphasizes the effectiveness and efficiency of\nthe proposed IDS.\n", " In most multiple-input multiple-output (MIMO) communication systems, e.g.,\nMassive MIMO, the antenna spacing is generally no less than half a wavelength.\nIt helps to reduce the mutual coupling and therefore facilitate the system\ndesign. The maximum array gain is the number of antennas in this settings.\nHowever, when the antenna spacing is made very small, the array gain of a\ncompact array can be proportional to the square of the number of antennas - a\nvalue much larger than the traditional array. To achieve this so-called\n\"superdirectivity\" however, the calculation of the excitation coefficients\n(beamforming vector) is known to be a challenging problem. In this paper, we\nderive the beamforming vector of superdirective arrays based on a novel\ncoupling matrix-enabled method. We also propose an approach to obtain the\ncoupling matrix, which is derived by the spherical wave expansion method and\nactive element pattern. The full-wave electromagnetic simulations are conducted\nto validate the effectiveness of our proposed method. Simulation results show\nthat when the beamforming vector obtained by our method is applied, the\ndirectivity of the designed dipole antenna array has a good agreement with the\ntheoretical values.\n", " Living cells can leverage correlations in environmental fluctuations to\npredict the future environment and mount a response ahead of time. To this end,\ncells need to encode the past signal into the output of the intracellular\nnetwork from which the future input is predicted. Yet, storing information is\ncostly while not all features of the past signal are equally informative on the\nfuture input signal. Here, we show, for two classes of input signals, that\ncellular networks can reach the fundamental bound on the predictive information\nas set by the information extracted from the past signal: push-pull networks\ncan reach this information bound for Markovian signals, while networks that\ntake a temporal derivative can reach the bound for predicting the future\nderivative of non-Markovian signals. However, the bits of past information that\nare most informative about the future signal are also prohibitively costly. As\na result, the optimal system that maximizes the predictive information for a\ngiven resource cost is, in general, not at the information bound. Applying our\ntheory to the chemotaxis network of Escherichia coli reveals that its adaptive\nkernel is optimal for predicting future concentration changes over a broad\nrange of background concentrations, and that the system has been tailored to\npredicting these changes in shallow gradients.\n", " In this paper, calculus of variation methods are generalized to find min-max\noptimal solution of uncertain dynamical systems with uncertain or certain cost.\nFirst, a new form of Euler-Lagrange conditions for uncertain systems is\npresented. Then several cases are indicated where final condition can be\nspecified or free. Also necessary conditions are introduced to existence of\nmin-max optimal solution of the uncertain systems. Finally, efficiency of the\nproposed method is verified through some examples.\n", " We unveil an alluring alternative to parametric search that applies to both\nthe non-geodesic and geodesic Fr\\'echet optimization problems. This randomized\napproach is based on a variant of red-blue intersections and is appealing due\nto its elegance and practical efficiency when compared to parametric search. We\npresent the first algorithm for the geodesic Fr\\'echet distance between two\npolygonal curves $A$ and $B$ inside a simple bounding polygon $P$. The geodesic\nFr\\'echet decision problem is solved almost as fast as its non-geodesic sibling\nand requires $O(N^{2\\log k)$ time and $O(k+N)$ space after $O(k)$\npreprocessing, where $N$ is the larger of the complexities of $A$ and $B$ and\n$k$ is the complexity of $P$. The geodesic Fr\\'echet optimization problem is\nsolved by a randomized approach in $O(k+N^{2\\log kN\\log N)$ expected time and\n$O(k+N^{2)$ space. This runtime is only a logarithmic factor larger than the\nstandard non-geodesic Fr\\'echet algorithm (Alt and Godau 1995). Results are\nalso presented for the geodesic Fr\\'echet distance in a polygonal domain with\nobstacles and the geodesic Hausdorff distance for sets of points or sets of\nline segments inside a simple polygon $P$.\n", " Relationship among efficiency and output power of heat-electric energy\nconverters as well as of any converters for transforming of heat energy into\nany other kind of energy is considered. It is shown, that the parameter\nefficiency does not determine univocally the output power of a converter. It is\nproposed to use another parameter for determination of working ability of heat\nenergy converters. It is shown, that high output power can not be achieved by\nany kind of Stirling-type converters in spite of their high efficiency.\n", " An accurate and efficient forecasting system is imperative to the prevention\nof emerging infectious diseases such as COVID-19 in public health. This system\nrequires accurate transient modeling, lower computation cost, and fewer\nobservation data. To tackle these three challenges, we propose a novel deep\nlearning approach using black-box knowledge distillation for both accurate and\nefficient transmission dynamics prediction in a practical manner. First, we\nleverage mixture models to develop an accurate, comprehensive, yet impractical\nsimulation system. Next, we use simulated observation sequences to query the\nsimulation system to retrieve simulated projection sequences as knowledge.\nThen, with the obtained query data, sequence mixup is proposed to improve query\nefficiency, increase knowledge diversity, and boost distillation model\naccuracy. Finally, we train a student deep neural network with the retrieved\nand mixed observation-projection sequences for practical use. The case study on\nCOVID-19 justifies that our approach accurately projects infections with much\nlower computation cost when observation data are limited.\n", " We study a genetic regulatory network model developed to demonstrate that\ngenetic robustness can evolve through stabilizing selection for optimal\nphenotypes. We report preliminary results on whether such selection could\nresult in a reorganization of the state space of the system. For the chosen\nparameters, the evolution moves the system slightly toward the more ordered\npart of the phase diagram. We also find that strong memory effects cause the\nDerrida annealed approximation to give erroneous predictions about the model's\nphase diagram.\n", " We consider a combined system of regular delivery trucks and crowdsourced\ndrones, available via a sharing economy platform, to provide a\ntechnology-assisted crowd-based last-mile delivery experience. We develop\nanalytical models and methods for a system in which package delivery is\nperformed by a big truck carrying many packages to a neighborhood or a town in\na metropolitan area and then the packages are assigned to crowdsourced drone\noperators to deliver them to their final destinations. We develop several\noptimization models for various cases of the problem and use a combination of\nheuristic algorithms to solve this NP-hard problem. Finally, we present\ncomputational results for the models and algorithms, and conduct an exhaustive\nsensitivity analysis to check the influence of different parameters and\nassumptions on the final results. We also provide extensive managerial insights\non the benefits of our proposed model if implemented in practice.\n", " Accelerated discovery with machine learning (ML) has begun to provide the\nadvances in efficiency needed to overcome the combinatorial challenge of\ncomputational materials design. Nevertheless, ML-accelerated discovery both\ninherits the biases of training data derived from density functional theory\n(DFT) and leads to many attempted calculations that are doomed to fail. Many\ncompelling functional materials and catalytic processes involve strained\nchemical bonds, open-shell radicals and diradicals, or metal-organic bonds to\nopen-shell transition-metal centers. Although promising targets, these\nmaterials present unique challenges for electronic structure methods and\ncombinatorial challenges for their discovery. In this Perspective, we describe\nthe advances needed in accuracy, efficiency, and approach beyond what is\ntypical in conventional DFT-based ML workflows. These challenges have begun to\nbe addressed through ML models trained to predict the results of multiple\nmethods or the differences between them, enabling quantitative sensitivity\nanalysis. For DFT to be trusted for a given data point in a high-throughput\nscreen, it must pass a series of tests. ML models that predict the likelihood\nof calculation success and detect the presence of strong correlation will\nenable rapid diagnoses and adaptation strategies. These \"decision engines\"\nrepresent the first steps toward autonomous workflows that avoid the need for\nexpert determination of the robustness of DFT-based materials discoveries.\n", " The modular design of planar phased arrays arranged on orthogonal\npolygon-shaped apertures is addressed and a new method is proposed to\nsynthesize domino-tiled arrays fitting multiple, generally conflicting,\nrequirements. Starting from an analytic procedure to check the\ndomino-tileability of the aperture, two multi-objective optimization techniques\nare derived to efficiently and effectively deal with small and medium/large\narrays depending on the values of the bounds for the cardinality of the\nsolution space of the admissible clustered solutions. A set of representative\nnumerical examples is reported to assess the effectiveness of the proposed\nsynthesis approach also through full-wave simulations when considering\nnon-ideal models for the radiating elements of the array.\n", " Modern Neural Network (NN) architectures heavily rely on vast numbers of\nmultiply-accumulate arithmetic operations, constituting the predominant\ncomputational cost. Therefore, this paper proposes a high-throughput, scalable\nand energy efficient non-element-wise matrix multiplication unit on FPGAs as a\nbasic component of the NNs. We firstly streamline inter-layer and intra-layer\nredundancies of MADDNESS algorithm, a LUT-based approximate matrix\nmultiplication, to design a fast, efficient scalable approximate matrix\nmultiplication module termed \"Approximate Multiplication Unit (AMU)\". The AMU\noptimizes LUT-based matrix multiplications further through dedicated memory\nmanagement and access design, decoupling computational overhead from input\nresolution and boosting FPGA-based NN accelerator efficiency significantly. The\nexperimental results show that using our AMU achieves up to 9x higher\nthroughput and 112x higher energy efficiency over the state-of-the-art\nsolutions for the FPGA-based Quantised Neural Network (QNN) accelerators.\n", " In recent years, several algorithms for system identification with neural\nstate-space models have been introduced. Most of the proposed approaches are\naimed at reducing the computational complexity of the learning problem, by\nsplitting the optimization over short sub-sequences extracted from a longer\ntraining dataset. Different sequences are then processed simultaneously within\na minibatch, taking advantage of modern parallel hardware for deep learning. An\nissue arising in these methods is the need to assign an initial state for each\nof the sub-sequences, which is required to run simulations and thus to evaluate\nthe fitting loss. In this paper, we provide insights for calibration of neural\nstate-space training algorithms based on extensive experimentation and analyses\nperformed on two recognized system identification benchmarks. Particular focus\nis given to the choice and the role of the initial state estimation. We\ndemonstrate that advanced initial state estimation techniques are really\nrequired to achieve high performance on certain classes of dynamical systems,\nwhile for asymptotically stable ones basic procedures such as zero or random\ninitialization already yield competitive performance.\n", " Inherent asymmetry in the tunneling barriers of few-electron quantum dots\ninduces intrinsically different tunneling currents for forward and reverse\nsource-drain biases in the non-linear transport regime. Here we show that in\naddition to spin selection rules, overlap matrix elements between many-body\nstates are crucial for the correct description of tunneling transmission\nthrough quantum dots at large magnetic fields. Signatures of excited\n(N-1)-electron states in the transport process through the N-electron system\nare clearly identified in the measured transconductances. Our analysis clearly\nconfirms the validity of single-electron quantum transport theory in quantum\ndots.\n", " Cells generally convert external nutrient resources to support metabolismand\ngrowth. Understanding the thermodynamic efficiency of this conversion is\nessential to determine the general characteristics of cellular growth. Using a\nsimple protocell model with catalytic reaction dynamics to synthesize the\nnecessary enzyme and membrane components from nutrients, the entropy production\nper unit cell-volume growth is calculated analytically and numerically based on\nthe rate equation for chemical kinetics and linear non-equilibrium\nthermodynamics. The minimal entropy production per unit cell growth is found to\nbe achieved at a non-zero nutrient uptake rate, rather than at a quasi-static\nlimit as in the standard Carnot engine. This difference appears because the\nequilibration mediated by the enzyme exists only within cells that grow through\nenzyme and membrane synthesis. Optimal nutrient uptake is also confirmed by\nprotocell models with many chemical components synthesized through a catalytic\nreaction network. The possible relevance of the identified optimal uptake to\noptimal yield for cellular growth is also discussed.\n", " A conversational recommender system (CRS) is a practical application for item\nrecommendation through natural language conversation. Such a system estimates\nuser interests for appropriate personalized recommendations. Users sometimes\nhave various interests in different categories or genres, but existing studies\nassume a unique user interest that can be covered by closely related items. In\nthis work, we propose to model such multiple user interests in CRS. We\ninvestigated its effects in experiments using the ReDial dataset and found that\nthe proposed method can recommend a wider variety of items than that of the\nbaseline CR-Walker.\n", " The validity of the Second Law of thermodynamics, indisputable in the\nmacroscopic world, is challenged at the mesoscopic level: a mesoscopic isolated\nsystem, possessing spatial dimensions of the order of a few microns, is\ncapable, as shown by a straightforward kinetic analysis, to exhibit a perpetuum\nmobile behavior associated with large negative variations of the Clausius\nentropy of the system. This violation of the Second Law is expedient for\ndevising a cyclic process through which an isolated system can extract energy\nfrom a surrounding thermal bath.\n", " We consider a processor sharing storage allocation model, which has m primary\nholding spaces and infinitely many secondary ones, and a single processor\nservicing the stored items (customers). All of the spaces are numbered and\nordered. An arriving customer takes the lowest available space. We define the\ntraffic intensity rho to be lambda/mu where lambda is the customers' arrival\nrate and mu is the service rate of the processor. We study the joint\nprobability distribution of the numbers of occupied primary and secondary\nspaces. For 0 < rho < 1, we obtain the exact solutions for m = 1 and m = 2. For\narbitrary m we study the problem in the asymptotic limit rho -> 1 with m fixed.\nWe also develop a semi-numerical semi-analytic method for computing the joint\ndistribution.\n", " In this paper, we consider the problem of partitioning a polygon into a set\nof connected disjoint sub-polygons, each of which covers an area of a specific\nsize. The work is motivated by terrain covering applications in robotics, where\nthe goal is to find a set of efficient plans for a team of heterogeneous robots\nto cover a given area. Within this application, solving a polygon partitioning\nproblem is an essential stepping stone. Unlike previous work, the problem\nformulation proposed in this paper also considers a compactness metric of the\ngenerated sub-polygons, in addition to the area size constraints. Maximizing\nthe compactness of sub-polygons directly influences the optimality of any\ngenerated motion plans. Consequently, this increases the efficiency with which\nrobotic tasks can be performed within each sub-region. The proposed problem\nrepresentation is based on grid cell decomposition and a potential field model\nthat allows for the use of standard optimization techniques. A new algorithm,\nthe AreaDecompose algorithm, is proposed to solve this problem. The algorithm\nincludes a number of existing and new optimization techniques combined with two\npost-processing methods. The approach has been evaluated on a set of randomly\ngenerated polygons which are then divided using different criteria and the\nresults have been compared with a state-of-the-art algorithm. Results show that\nthe proposed algorithm can efficiently divide polygon regions maximizing\ncompactness of the resulting partitions, where the sub-polygon regions are on\naverage up to 73% more compact in comparison to existing techniques.\n", " We explore efficient optimization of toolpaths based on multiple criteria for\nlarge instances of 3D printing problems. We first show that the minimum turn\ncost 3D printing problem is NP-hard, even when the region is a simple polygon.\nWe develop SFCDecomp, a space filling curve based decomposition framework to\nsolve large instances of 3D printing problems efficiently by solving these\noptimization subproblems independently. For the Buddha model, our framework\nbuilds toolpaths over a total of 799,716 nodes across 169 layers, and for the\nBunny model it builds toolpaths over 812,733 nodes across 360 layers. Building\non SFCDecomp, we develop a multicriteria optimization approach for toolpath\nplanning. We demonstrate the utility of our framework by maximizing or\nminimizing tool path edge overlap between adjacent layers, while jointly\nminimizing turn costs. Strength testing of a tensile test specimen printed with\ntool paths that maximize or minimize adjacent layer edge overlaps reveal\nsignificant differences in tensile strength between the two classes of prints.\n", " Geometry processing presents a variety of difficult numerical problems, each\nseeming to require its own tailored solution. This breadth is largely due to\nthe expansive list of geometric primitives, e.g., splines, triangles, and\nhexahedra, joined with an ever-expanding variety of objectives one might want\nto achieve with them. With the recent increase in attention toward higher-order\nsurfaces, we can expect a variety of challenges porting existing solutions that\nwork on triangle meshes to work on these more complex geometry types. In this\npaper, we present a framework for solving many core geometry processing\nproblems on higher-order surfaces. We achieve this goal through sum-of-squares\noptimization, which transforms nonlinear polynomial optimization problems into\nsequences of convex problems whose complexity is captured by a single degree\nparameter. This allows us to solve a suite of problems on higher-order\nsurfaces, such as continuous collision detection and closest point queries on\ncurved patches, with only minor changes between formulations and geometries.\n" ]

[ " An alternative proof is given of the existence of greatest lower bounds in\nthe imbalance order of binary maximal instantaneous codes of a given size.\nThese codes are viewed as maximal antichains of a given size in the infinite\nbinary tree of 0-1 words. The proof proposed makes use of a single balancing\noperation instead of expansion and contraction as in the original proof of the\nexistence of glb.\n", " The aim of this work is to give degree formulas for the generalized Hamming\nweights of evaluation codes and to show lower bounds for these weights. In\nparticular, we give degree formulas for the generalized Hamming weights of\nReed--Muller-type codes, and we determine the minimum distance of toric codes\nover hypersimplices, and the 1st and 2nd generalized Hamming weights of\nsquarefree evaluation codes.\n", " Huffman Codes are optimal Instantaneous Fixed-to-Variable (FV) codes in which\nevery source symbol can only be encoded by one codeword. Relaxing these\nconstraints permits constructing better FV codes. More specifically, recent\nwork has shown that AIFV-$m$ codes can beat Huffman coding. AIFV-$m$ codes\nconstruct am $m$-tuple of different coding trees between which the code\nalternates and are only almost instantaneous (AI). This means that decoding a\nword might require a delay of a finite number of bits.\n Current algorithms for constructing optimal AIFV-$m$ codes are iterative\nprocesses that construct progressively \"better sets\" of code trees. The\nprocesses have been proven to finitely converge to the optimal code but with no\nknown bounds on the convergence rate.\n This paper derives a geometric interpretation of the space of binary AIFV-$2$\ncodes, permitting the development of the first polynomially time-bounded\niterative procedures for constructing optimal AIFV codes.\n We first show that a simple binary search procedure can replace the current\niterative process to construct optimal binary AIFV-$2$ codes. We then describe\nhow to frame the problem as a linear programming with an exponential number of\nconstraints but a polynomial-time separability oracle. This permits using the\nGr\\\"otschel, Lov\\'asz and Schrijver ellipsoid method to solve the problem in a\npolynomial number of steps. While more complicated, this second method has the\npotential to lead to a polynomial time algorithm to construct optimal AIFV-$m$\ncodes for general $m$.\n", " We discuss, in terms of rate-distortion theory, the fitness of molecular\ncodes as the problem of designing an optimal information channel. The fitness\nis governed by an interplay between the cost and quality of the channel, which\ninduces smoothness in the code. By incorporating this code fitness into\npopulation dynamics models, we suggest that the emergence and evolution of\nmolecular codes may be explained by simple channel design considerations.\n", " To address growth challenges facing large Data Centers and supercomputing\nclusters a new construction is presented for scalable, high throughput, low\nlatency networks. The resulting networks require 1.5-5 times fewer switches,\n2-6 times fewer cables, have 1.2-2 times lower latency and correspondingly\nlower congestion and packet losses than the best present or proposed networks\nproviding the same number of ports at the same total bisection. These advantage\nratios increase with network size. The key new ingredient is the exact\nequivalence discovered between the problem of maximizing network bisection for\nlarge classes of practically interesting Cayley graphs and the problem of\nmaximizing codeword distance for linear error correcting codes. Resulting\ntranslation recipe converts existent optimal error correcting codes into\noptimal throughput networks.\n", " Convolutional codes are constructed, designed and analysed using row and/or\nblock structures of unit algebraic schemes. Infinite series of such codes and\nof codes with specific properties are derived. Properties are shown\nalgebraically and algebraic decoding methods are derived. For a given rate and\ngiven error-correction capability at each component, convolutional codes with\nthese specifications and with efficient decoding algorithms are constructed.\nExplicit prototype examples are given but in general large lengths and large\nerror capability are achievable. Convolutional codes with efficient decoding\nalgorithms at or near the maximum free distances attainable for the parameters\nare constructible. Unit memory convolutional codes of maximum possible free\ndistance are designed with practical algebraic decoding algorithms.\n LDPC (low density parity check) convolutional codes with efficient decoding\nschemes are constructed and analysed by the methods. Self-dual and\ndual-containing convolutional codes may also be designed by the methods;\ndual-containing codes enables the construction of quantum codes.\n", " Neural codes serve as a language for neurons in the brain. Convex codes,\nwhich arise from the pattern of intersections of convex sets in Euclidean\nspace, are of particular relevance to neuroscience. Not every code is convex,\nhowever, and the combinatorial properties of a code that determine its\nconvexity are still poorly understood. Here we find that a code that can be\nrealized by a collection of open convex sets may or may not be realizable by\nclosed convex sets, and vice versa, establishing that open convex and closed\nconvex codes are distinct classes. We also prove that max intersection-complete\ncodes (i.e. codes that contain all intersections of maximal codewords) are both\nopen convex and closed convex, and provide an upper bound for their minimal\nembedding dimension. Finally, we show that the addition of non-maximal\ncodewords to an open convex code preserves convexity.\n", " We consider the problem of computing the Fr\\'echet distance between two\ncurves for which the exact locations of the vertices are unknown. Each vertex\nmay be placed in a given uncertainty region for that vertex, and the objective\nis to place vertices so as to minimise the Fr\\'echet distance. This problem was\nrecently shown to be NP-hard in 2D, and it is unclear how to compute an optimal\nvertex placement at all.\n We present the first general algorithmic framework for this problem. We prove\nthat it results in a polynomial-time algorithm for curves in 1D with intervals\nas uncertainty regions. In contrast, we show that the problem is NP-hard in 1D\nin the case that vertices are placed to maximise the Fr\\'echet distance.\n We also study the weak Fr\\'echet distance between uncertain curves. While\nfinding the optimal placement of vertices seems more difficult than the regular\nFr\\'echet distance -- and indeed we can easily prove that the problem is\nNP-hard in 2D -- the optimal placement of vertices in 1D can be computed in\npolynomial time. Finally, we investigate the discrete weak Fr\\'echet distance,\nfor which, somewhat surprisingly, the problem is NP-hard already in 1D.\n", " We consider a mobile edge computing scenario where users want to perform a\nlinear inference operation $\\boldsymbol{W} \\boldsymbol{x}$ on local data\n$\\boldsymbol{x}$ for some network-side matrix $\\boldsymbol{W}$. The inference\nis performed in a distributed fashion over multiple servers at the network\nedge. For this scenario, we propose a coding scheme that combines a rateless\ncode to provide resiliency against straggling servers--hence reducing the\ncomputation latency--and an irregular-repetition code to provide spatial\ndiversity--hence reducing the communication latency. We further derive a lower\nbound on the total latency--comprising computation latency, communication\nlatency, and decoding latency. The proposed scheme performs remarkably close to\nthe bound and yields significantly lower latency than the scheme based on\nmaximum distance separable codes recently proposed by Zhang and Simeone.\n", " Let $\\mathcal{C}$ be a set of $m$ by $n$ matrices over $\\mathbb{F}_q$ such\nthat the rank of $A-B$ is at least $d$ for all distinct $A,B\\in \\mathcal{C}$.\nSuppose that $m\\leqslant n$. If $\\#\\mathcal{C}= q^{n(m-d+1)}$, then\n$\\mathcal{C}$ is a maximum rank distance (MRD for short) code. Until 2016,\nthere were only two known constructions of MRD codes for arbitrary $1<d<m-1$.\nOne was found by Delsarte (1978) and Gabidulin (1985) independently, and it was\nlater generalized by Kshevetskiy and Gabidulin (2005). We often call them\n(generalized) Gabidulin codes. Another family was recently obtained by Sheekey\n(2016), and its elements are called twisted Gabidulin codes. In the same paper,\nSheekey also proposed a generalization of the twisted Gabidulin codes. However\nthe equivalence problem for it is not considered, whence it is not clear\nwhether there exist new MRD codes in this generalization. We call the members\nof this putative larger family generalized twisted Gabidulin codes. In this\npaper, we first compute the Delsarte duals and adjoint codes of them, then we\ncompletely determine the equivalence between different generalized twisted\nGabidulin codes. In particular, it can be proven that, up to equivalence,\ngeneralized Gabidulin codes and twisted Gabidulin codes are both proper subsets\nof this family.\n", " Starting with a similarity function between objects, it is possible to define\na distance metric on pairs of objects, and more generally on probability\ndistributions over them. These distance metrics have a deep basis in functional\nanalysis, measure theory and geometric measure theory, and have a rich\nstructure that includes an isometric embedding into a (possibly infinite\ndimensional) Hilbert space. They have recently been applied to numerous\nproblems in machine learning and shape analysis.\n In this paper, we provide the first algorithmic analysis of these distance\nmetrics. Our main contributions are as follows: (i) We present fast\napproximation algorithms for computing the kernel distance between two point\nsets P and Q that runs in near-linear time in the size of (P cup Q) (note that\nan explicit calculation would take quadratic time). (ii) We present\npolynomial-time algorithms for approximately minimizing the kernel distance\nunder rigid transformation; they run in time O(n + poly(1/epsilon, log n)).\n(iii) We provide several general techniques for reducing complex objects to\nconvenient sparse representations (specifically to point sets or sets of points\nsets) which approximately preserve the kernel distance. In particular, this\nallows us to reduce problems of computing the kernel distance between various\ntypes of objects such as curves, surfaces, and distributions to computing the\nkernel distance between point sets. These take advantage of the reproducing\nkernel Hilbert space and a new relation linking binary range spaces to\ncontinuous range spaces with bounded fat-shattering dimension.\n", " We discuss the continuum field theory limit of the physical scenario\ndescribed in Ref. [1], the universe arising from the interpretation of the most\ngeneral collection of logical codes in terms of distributions of units of\nenergy along units of space. This limit leads in a natural way to string theory\nas the theory which allows to perturbatively parametrize the geometric\nstructures in terms of propagating particles and fields. We discuss some\ngeneral properties of the spectrum, masses and couplings, the existence of the\nstrong force, with particular attention to the excited states, and the\nimplications for the physics of high energy colliders.\n", " In this article we present a class of codes with few weights arising from\nspecial type of linear sets. We explicitly show the weights of such codes,\ntheir weight enumerator and possible choices for their generator matrices. In\nparticular, our construction yields also to linear codes with three weights\nand, in some cases, to almost MDS codes. The interest for these codes relies on\ntheir applications to authentication codes and secret schemes, and their\nconnections with further objects such as association schemes and graphs.\n", " We characterize the semigroups of composition operators that are strongly\ncontinuous on the mixed norm spaces $H(p,q,\\alpha)$. First, we study the\nseparable spaces $H(p,q,\\alpha)$ with $q<\\infty,$ that behave as the Hardy and\nBergman spaces. In the second part we deal with the spaces\n$H(p,\\infty,\\alpha),$ where polynomials are not dense. To undertake this study,\nwe introduce the integral operators, characterize its boundedness and\ncompactness, and use its properties to find the maximal closed linear subspace\nof $H(p,\\infty,\\alpha)$ in which the semigroups are strongly continuous. In\nparticular, we obtain that this maximal space is either $H(p,0,\\alpha)$ or\nnon-separable, being this result the deepest one in the paper.\n", " This paper describes a novel method of generating hex-dominant meshes using\npre-computed optimal subdivisions of the unit cube in a grid-based approach.\nOur method addresses geometries that are standard in mechanical engineering and\noften must comply with the restrictions of subtractive manufacturability. A\ncentral component of our method is the set of subdivisions we pre-compute with\nAnswer Set Programming. Despite being computationally expensive, we obtain\noptimal meshes of up to 35 nodes available to our method in a template fashion.\nThe first step in our grid-based method generates a coarse Precursor Mesh for\nmeshing complete parts representing the bar stock. Then, the resulting mesh is\ngenerated in a subtractive manner by inserting and fitting the pre-generated\nsubdivisions into the Precursor Mesh. This step guarantees that the elements\nare of good quality. In the final stage, the mesh nodes are mapped to geometric\nentities of the target geometry to get an exact match. We demonstrate our\nmethod with multiple examples showing the strength of this approach.\n", " We construct the spin flaglet transform, a wavelet transform to analyze spin\nsignals in three dimensions. Spin flaglets can probe signal content localized\nsimultaneously in space and frequency and, moreover, are separable so that\ntheir angular and radial properties can be controlled independently. They are\nparticularly suited to analyzing of cosmological observations such as the weak\ngravitational lensing of galaxies. Such observations have a unique 3D\ngeometrical setting since they are natively made on the sky, have spin angular\nsymmetries, and are extended in the radial direction by additional distance or\nredshift information. Flaglets are constructed in the harmonic space defined by\nthe Fourier-Laguerre transform, previously defined for scalar functions and\nextended here to signals with spin symmetries. Thanks to various sampling\ntheorems, both the Fourier-Laguerre and flaglet transforms are theoretically\nexact when applied to bandlimited signals. In other words, in numerical\ncomputations the only loss of information is due to the finite representation\nof floating point numbers. We develop a 3D framework relating the weak lensing\npower spectrum to covariances of flaglet coefficients. We suggest that the\nresulting novel flaglet weak lensing estimator offers a powerful alternative to\ncommon 2D and 3D approaches to accurately capture cosmological information.\nWhile standard weak lensing analyses focus on either real or harmonic space\nrepresentations (i.e., correlation functions or Fourier-Bessel power spectra,\nrespectively), a wavelet approach inherits the advantages of both techniques,\nwhere both complicated sky coverage and uncertainties associated with the\nphysical modeling of small scales can be handled effectively. Our codes to\ncompute the Fourier-Laguerre and flaglet transforms are made publicly\navailable.\n", " End-point maximal $L^1$-regularity for parabolic initial-boundary value\nproblems is considered. For the inhomogeneous Dirichlet and Neumann data,\nmaximal $L^1$-regularity for initial-boundary value problems is established in\ntime end-point case upon the homogeneous Besov space $\\dot B_{p,1}^s({\\mathbb\nR}^n_+)$ with $1< p< \\infty$ and $-1+1/p<s\\le 0$ as well as optimal trace\nestimates. The main estimates obtained here are sharp in the sense of trace\nestimates and it is not available by known theory on the class of UMD Banach\nspaces. We utilize a method of harmonic analysis, in particular, the almost\northogonal properties between the boundary potentials of the Dirichlet and the\nNeumann boundary data and the Littlewood-Paley dyadic decomposition of unity in\nthe Besov and the Lizorkin-Triebel spaces.\n", " We construct a family of n+1 dyadic filtrations in R^n, so that every\nEuclidean ball B is contained in some cube Q of our family satisfying diam(Q)\n\\le c_n diam(B) for some dimensional constant c_n. Our dyadic covering is\noptimal on the number of filtrations and improves previous results of Christ\nand Garnett/Jones by extending a construction of Mei for the n-torus. Based on\nthis covering and motivated by applications to matrix-valued functions, we\nprovide a dyadic nondoubling Calder\\'on-Zygmund decomposition which avoids\nBesicovitch type coverings in Tolsa's decomposition. We also use a recent\nresult of Hyt\\\"onen and Kairema to extend our dyadic nondoubling decomposition\nto the more general setting of upper doubling metric spaces.\n", " In this article, we introduce and study Riesz bases in a separable\nquaternionic Hilbert spaces. Some results on Riesz bases in a separable\nquaternionic Hilbert spaces are proved. It is also proved that a Riesz basis in\na separable quaternionic Hilbert space a frame for the quaternionic Hilbert\nspace. Riesz sequences are defined and equivalence of a Riesz basis and a\ncomplete Riesz sequence in a separable quaternionic Hilbert space is proved.\n", " A line packing is optimal if its coherence is as small as possible. Most\ninteresting examples of optimal line packings are achieving equality in some of\nthe known lower bounds for coherence. In this paper two infinite families of\nreal and complex biangular line packings are presented. New packings achieve\nequality in the real or complex second Levenshtein bound respectively. Both\ninfinite families are constructed by analyzing well known representations of\nthe finite groups SL$(2,\\mathbb{F}_q)$. Until now the only known infinite\nfamiles meeting the second Levenshtein bounds were related to the maximal sets\nof mutually unbiased bases (MUB). Similarly to the line packings related to the\nmaximal sets of MUBs, the line packings presented here are related to the\nmaximal sets of mutually unbiased weighing matrices. Another similarity is that\nthe new packings are projective 2-designs. The latter property together with\nsufficiently large cardinalities of the new packings implies some improvement\non largest known cardinalities of real and complex biangular tight frames.\n", " For Schr\\\"{o}dinger type operators in one dimension, we consider the\nrelationship between the convergence rate and the regularity for initial data.\nBy establishing the associated frequency-localized maximal estimates, we prove\nsharp results up to the endpoints. The optimal range for the wave operator in\nall dimensions is also obtained.\n", " In this paper, we define Abelian and consta-Abelian polyadic codes over rings\ndefined as affine algebras over chain rings. For that aim, we use the classical\nconstruction via splittings and multipliers of the underlying Abelian group. We\nalso derive some results on the structure of the associated polyadic codes and\nthe number of codes under these conditions.\n", " In this paper we analyse the construction of identification codes.\nIdentification codes are based on the question \"Is the message I have just\nreceived the one I am interested in?\", as opposed to Shannon's transmission,\nwhere the receiver is interested in not only one, but any message. The\nadvantage of identification is that it allows rates growing double\nexponentially in the blocklength at the cost of not being able to decode every\nmessage, which might be beneficial in certain applications. We focus on a\nspecial identification code construction based on two concatenated Reed-Solomon\ncodes and have a closer look at its implementation, analyzing the trade-offs of\nidentification with respect to transmission and the trade-offs introduced by\nthe computational cost of identification codes.\n", " Given two point sets in the plane, we study the minimization of the\nbottleneck distance between a point set B and an equally-sized subset of a\npoint set A under translations. We relate this problem to a Voronoi-type\ndiagram and derive polynomial bounds for its complexity that are optimal in the\nsize of A. We devise efficient algorithms for the construction of such a\ndiagram and its lexicographic variant, which generalize to higher dimensions.\nWe use the diagram to find an optimal bottleneck matching under translations,\nto compute a connecting path of minimum bottleneck cost between two positions\nof B, and to determine the maximum bottleneck cost in a convex polygon.\n", " The class of Bipartite Distance Hereditary (BDH) graphs is the intersection\nbetween bipartite domino-free and chordal bipartite graphs.\\ Graphs in both the\nlatter classes have linearly many maximal bicliques, implying the existence of\npolynomial-time algorithms for computing the associated Galois lattice.\\ Such a\nlattice can indeed be built in $O(m\\times n)$ worst case-time for a domino-free\ngraph with $m$ edges and $n$ vertices.\\ In this paper we give a sharp estimate\non the number of the maximal bicliques of BDH graphs and exploit such result to\ngive an $O(m)$ worst case time algorithm for computing the Galois lattice of\nBDH graphs. By relying on the fact that neighborhoods of vertices of BDH graphs\ncan be realized as directed paths in a arborescence, we give an $O(n)$\nworst-case space and time encoding of both the input graph and its Galois\nlattice, provided that the reverse of a Bandelt and Mulder building sequence is\ngiven.\n", " We investigate the autonomous navigation of a mobile robot in the presence of\nother moving vehicles under time-varying uncertain environmental disturbances.\nWe first predict the future state distributions of other vehicles to account\nfor their uncertain behaviors affected by the time-varying disturbances. We\nthen construct a dynamic-obstacle-aware reachable space that contains states\nwith high probabilities to be reached by the robot, within which the optimal\npolicy is searched. Since, in general, the dynamics of both the vehicle and the\nenvironmental disturbances are nonlinear, we utilize a nonlinear Gaussian\nfilter -- the unscented transform -- to approximate the future state\ndistributions. Finally, the forward reachable space computation and backward\npolicy search are iterated until convergence. Extensive simulation evaluations\nhave revealed significant advantages of this proposed method in terms of\ncomputation time, decision accuracy, and planning reliability.\n", " The exploration of linear subspaces, particularly scattered subspaces, has\ngarnered considerable attention across diverse mathematical disciplines in\nrecent years, notably within finite geometries and coding theory. Scattered\nsubspaces play a pivotal role in analyzing various geometric structures such as\nblocking sets, two-intersection sets, complete arcs, caps in affine and\nprojective spaces over finite fields and rank metric codes. This paper\nintroduces a new infinite family of $h$-subspaces, along with their associated\nMRD codes. Additionally, it addresses the task of determining the generalized\nweights of these codes. Notably, we demonstrate that these MRD codes exhibit\nsome larger generalized weights compared to those previously identified.\n", " Codes in finite projective spaces equipped with the subspace distance have\nbeen proposed for error control in random linear network coding. The resulting\nso-called \\emph{Main Problem of Subspace Coding} is to determine the maximum\nsize $A_q(v,d)$ of a code in $\\operatorname{PG}(v-1,\\mathbb{F}_q)$ with minimum\nsubspace distance $d$. Here we completely resolve this problem for $d\\ge v-1$.\nFor $d=v-2$ we present some improved bounds and determine $A_q(5,3)=2q^3+2$\n(all $q$), $A_2(7,5)=34$. We also provide an exposition of the known\ndetermination of $A_q(v,2)$, and a table with exact results and bounds for the\nnumbers $A_2(v,d)$, $v\\leq 7$.\n", " We introduce a new variant of the art gallery problem that comes from safety\nissues. In this variant we are not interested in guard sets of smallest\ncardinality, but in guard sets with largest possible distances between these\nguards. To the best of our knowledge, this variant has not been considered\nbefore. We call it the Dispersive Art Gallery Problem. In particular, in the\ndispersive art gallery problem we are given a polygon $\\mathcal{P}$ and a real\nnumber $\\ell$, and want to decide whether $\\mathcal{P}$ has a guard set such\nthat every pair of guards in this set is at least a distance of $\\ell$ apart.\n In this paper, we study the vertex guard variant of this problem for the\nclass of polyominoes. We consider rectangular visibility and distances as\ngeodesics in the $L_1$-metric. Our results are as follows. We give a (simple)\nthin polyomino such that every guard set has minimum pairwise distances of at\nmost $3$. On the positive side, we describe an algorithm that computes guard\nsets for simple polyominoes that match this upper bound, i.e., the algorithm\nconstructs worst-case optimal solutions. We also study the computational\ncomplexity of computing guard sets that maximize the smallest distance between\nall pairs of guards within the guard sets. We prove that deciding whether there\nexists a guard set realizing a minimum pairwise distance for all pairs of\nguards of at least $5$ in a given polyomino is NP-complete. We were also able\nto find an optimal dynamic programming approach that computes a guard set that\nmaximizes the minimum pairwise distance between guards in tree-shaped\npolyominoes, i.e., computes optimal solutions. Because the shapes constructed\nin the NP-hardness reduction are thin as well (but have holes), this result\ncompletes the case for thin polyominoes.\n", " Networks of neurons in the brain encode preferred patterns of neural activity\nvia their synaptic connections. Despite receiving considerable attention, the\nprecise relationship between network connectivity and encoded patterns is still\npoorly understood. Here we consider this problem for networks of\nthreshold-linear neurons whose computational function is to learn and store a\nset of binary patterns (e.g., a neural code) as \"permitted sets\" of the\nnetwork. We introduce a simple Encoding Rule that selectively turns \"on\"\nsynapses between neurons that co-appear in one or more patterns. The rule uses\nsynapses that are binary, in the sense of having only two states (\"on\" or\n\"off\"), but also heterogeneous, with weights drawn from an underlying synaptic\nstrength matrix S. Our main results precisely describe the stored patterns that\nresult from the Encoding Rule -- including unintended \"spurious\" states -- and\ngive an explicit characterization of the dependence on S. In particular, we\nfind that binary patterns are successfully stored in these networks when the\nexcitatory connections between neurons are geometrically balanced -- i.e., they\nsatisfy a set of geometric constraints. Furthermore, we find that certain types\nof neural codes are \"natural\" in the context of these networks, meaning that\nthe full code can be accurately learned from a highly undersampled set of\npatterns. Interestingly, many commonly observed neural codes in cortical and\nhippocampal areas are natural in this sense. As an application, we construct\nnetworks that encode hippocampal place field codes nearly exactly, following\npresentation of only a small fraction of patterns. To obtain our results, we\nprove new theorems using classical ideas from convex and distance geometry,\nsuch as Cayley-Menger determinants, revealing a novel connection between these\nareas of mathematics and coding properties of neural networks.\n", " In this paper, we continue the investigation of topological properties of\nunbounded norm (un-)topology in normed lattices. We characterize separability\nand second countability of un-topology in terms of properties of the underlying\nnormed lattice. We apply our results to prove that an order continuous Banach\nfunction space $X$ over a semi-finite measure space is separable if and only if\nit has a $\\sigma$-finite carrier and is separable with respect to the topology\nof local convergence in measure. We also address the question when a normed\nlattice is a normal space with respect to the un-topology.\n", " We present a new method based on functional tensor decomposition and dynamic\ntensor approximation to compute the solution of a high-dimensional\ntime-dependent nonlinear partial differential equation (PDE). The idea of\ndynamic approximation is to project the time derivative of the PDE solution\nonto the tangent space of a low-rank functional tensor manifold at each time.\nSuch a projection can be computed by minimizing a convex energy functional over\nthe tangent space. This minimization problem yields the unique optimal velocity\nvector that allows us to integrate the PDE forward in time on a tensor manifold\nof constant rank. In the case of initial/boundary value problems defined in\nreal separable Hilbert spaces, this procedure yields evolution equations for\nthe tensor modes in the form of a coupled system of one-dimensional\ntime-dependent PDEs. We apply the dynamic tensor approximation to a\nfour-dimensional Fokker-Planck equation with non-constant drift and diffusion\ncoefficients, and demonstrate its accuracy in predicting relaxation to\nstatistical equilibrium.\n", " We show that a minor variant of the continuous Fr\\'echet distance between\npolygonal curves can be computed using essentially the same algorithm used to\nsolve the discrete version, thus dramatically simplifying the algorithm for\ncomputing it. The new variant is not necessarily monotone, but this shortcoming\ncan be easily handled via refinement.\n Combined with a Dijkstra/Prim type algorithm, this leads to a realization of\nthe Fr\\'echet distance (i.e., a morphing) that is locally optimal (aka locally\ncorrect), that is both easy to compute, and in practice, takes near linear time\non many inputs. The new morphing has the property that the leash is always as\nshort-as-possible.\n We implemented the new algorithm, and developed various strategies to get a\nfast execution in practice. Among our new contributions is a new simplification\nstrategy that is distance-sensitive, and enables us to compute the exact\ncontinuous Fr\\'echet distance in near linear time in practice. We preformed\nextensive experiments on our new algorithm, and released \\texttt{Julia} and\n\\texttt{Python} packages with these new implementations.\n", " One of the main problems of the research area of network coding is to compute\ngood lower and upper bounds of the achievable cardinality of so-called subspace\ncodes in $\\operatorname{PG}(n,q)$, i.e., the set of subspaces of\n$\\mathbb{F}_q^n$, for a given minimal distance. Here we generalize a\nconstruction of Etzion and Silberstein to a wide range of parameters. This\nconstruction, named coset construction, improves or attains several of the\npreviously best-known subspace code sizes and attains the MRD bound for an\ninfinite family of parameters.\n", " Persistence diagrams have been widely used to quantify the underlying\nfeatures of filtered topological spaces in data visualization. In many\napplications, computing distances between diagrams is essential; however,\ncomputing these distances has been challenging due to the computational cost.\nIn this paper, we propose a persistence diagram hashing framework that learns a\nbinary code representation of persistence diagrams, which allows for fast\ncomputation of distances. This framework is built upon a generative adversarial\nnetwork (GAN) with a diagram distance loss function to steer the learning\nprocess. Instead of using standard representations, we hash diagrams into\nbinary codes, which have natural advantages in large-scale tasks. The training\nof this model is domain-oblivious in that it can be computed purely from\nsynthetic, randomly created diagrams. As a consequence, our proposed method is\ndirectly applicable to various datasets without the need for retraining the\nmodel. These binary codes, when compared using fast Hamming distance, better\nmaintain topological similarity properties between datasets than other\nvectorized representations. To evaluate this method, we apply our framework to\nthe problem of diagram clustering and we compare the quality and performance of\nour approach to the state-of-the-art. In addition, we show the scalability of\nour approach on a dataset with 10k persistence diagrams, which is not possible\nwith current techniques. Moreover, our experimental results demonstrate that\nour method is significantly faster with the potential of less memory usage,\nwhile retaining comparable or better quality comparisons.\n", " We seek to augment a geometric network in the Euclidean plane with shortcuts\nto minimize its continuous diameter, i.e., the largest network distance between\nany two points on the augmented network. Unlike in the discrete setting where a\nshortcut connects two vertices and the diameter is measured between vertices,\nwe take all points along the edges of the network into account when placing a\nshortcut and when measuring distances in the augmented network.\n We study this network augmentation problem for paths and cycles. For paths,\nwe determine an optimal shortcut in linear time. For cycles, we show that a\nsingle shortcut never decreases the continuous diameter and that two shortcuts\nalways suffice to reduce the continuous diameter. Furthermore, we characterize\noptimal pairs of shortcuts for convex and non-convex cycles. Finally, we\ndevelop a linear time algorithm that produces an optimal pair of shortcuts for\nconvex cycles. Apart from the algorithms, our results extend to rectifiable\ncurves.\n Our work reveals some of the underlying challenges that must be overcome when\naddressing the discrete version of this network augmentation problem, where we\nminimize the discrete diameter of a network with shortcuts that connect only\nvertices.\n", " N players are randomly fitted with a colored hat (q different colors). All\nplayers guess simultaneously the color of their own hat observing only the hat\ncolors of the other N-1 players. The team wins if all players guess right. No\ncommunication of any sort is allowed, except for an initial strategy session\nbefore the game begins. In the first part of our investigation we have q\ndifferent colors with equal probability. Up to 4 colors we construct optimal\nstrategies for any number of players using Hamming Complete Sets. For 5 colors\nwe find optimal strategies up to 5 players using Optimal Hamming Sets. In the\nsecond part we have two colors where the probabilities may differ. We construct\noptimal strategies and maximal probability of winning the game for any number\nof players.\n", " In this paper, we consider the Laplace operator on the half-space with\nDirichlet and Neumann boundary conditions. We prove that this operator admits a\nbounded $H^\\infty$-calculus on Sobolev spaces with power weights measuring the\ndistance to the boundary. These weights do not necessarily belong to the class\nof Muckenhoupt $A_p$ weights. We additionally study the corresponding Dirichlet\nand Neumann heat semigroup. It is shown that these semigroups, in contrast to\nthe $L^p$-case, have polynomial growth. Moreover, maximal regularity results\nfor the heat equation are derived on inhomogeneous and homogeneous weighted\nSobolev spaces.\n", " Given a n points in two dimensional space, a Manhattan Network G is a network\nthat connects all n points with either horizontal or vertical edges, with the\nproperty that for any two point in G should be connected by a Manhattan path\nand distance between this two points is equal to Manhattan Distance. The\nMinimum Manhattan Network problem is to find a Manhattan network with minimum\nnetwork length, i.e., summation of all line segment in network should be\nminimize. In this paper, we proposed a 2-approximation algorithm with time\ncomplexity O(|E|lgN) where |E| is the number of edges and N is the number of\nnodes. Using randomly generated datasets, we compare our result with the\noptimal one.\n", " A useful heuristic in the understanding of large random combinatorial\nstructures is the Arratia-Tavare principle, which describes an approximation to\nthe joint distribution of component-sizes using independent random variables.\nThe principle outlines conditions under which the total variation distance\nbetween the true joint distribution and the approximation should be small, and\nwas successfully exploited by Pittel in the cases of integer partitions and set\npartitions. We provide sufficient conditions for this principle to be true in a\ngeneral context, valid for certain discrete probability distributions which are\n$\\textit{perturbed log-concave}$, via a quantitative local central limit\ntheorem. We then use it to generalize some classical asymptotic statistics in\ncombinatorial theory, as well as assert some new ones.\n", " An incremental approach for computation of convex hull for data points in\ntwo-dimensions is presented. The algorithm is not output-sensitive and costs a\ntime that is linear in the size of data points at input. Graham's scan is\napplied only on a subset of the data points, represented at the extremal of the\ndataset. Points are classified for extremal, in proportion with the modular\ndistance, about an imaginary point interior to the region bounded by convex\nhull of the dataset assumed for origin or center in polar coordinate. A subset\nof the data is arrived by terminating at until an event of no change in maximal\npoints is observed per bin, for iteratively and exponentially decreasing\nintervals.\n", " Two planar sets are circularly separable if there exists a circle enclosing\none of the sets and whose open interior disk does not intersect the other set.\n This paper studies two problems related to circular separability. A\nlinear-time algorithm is proposed to decide if two polygons are circularly\nseparable. The algorithm outputs the smallest separating circle. The second\nproblem asks for the largest circle included in a preprocessed, convex polygon,\nunder some point and/or line constraints. The resulting circle must contain the\nquery points and it must lie in the halfplanes delimited by the query lines.\n", " The epsilon-enlargement of a maximal monotone operator is a construct similar\nto the Br{\\o}ndsted and Rocakfellar epsilon-subdifferential enlargement of the\nsubdifferential. Like the epsilon-subdifferential, the epsilon-enlargement of a\nmaximal monotone operator has practical and theoretical applications.\n In a recent paper in Journal of Convex Analysis Burachik and Iusem studied\nconditions under which a maximal monotone operator is non-enlargeable, that is,\nits epsilon-enlargement coincides with the operator. Burachik and Iusem studied\nthese non-enlargeable operators in reflexive Banach spaces, assuming the\ninterior of the domain of the operator to be nonempty. In the present work, we\nremove the assumption on the domain of non-enlargeable operators and also\npresent partial results for the non-reflexive case.\n", " We show that the problem of computing the distance of a given permutation\nfrom a subgroup $H$ of $S_n$ is in general NP-complete, even under the\nrestriction that $H$ is elementary Abelian of exponent 2. The problem is shown\nto be polynomial-time equivalent to a problem related to finding a maximal\npartition of the edges of an Eulerian directed graph into cycles and this\nproblem is in turn equivalent to the standard NP-complete problem of Boolean\nsatisfiability.\n", " In nonadaptive group testing, the main research objective is to design an\nefficient algorithm to identify a set of up to $t$ positive elements among $n$\nsamples with as few tests as possible. Disjunct matrices and separable matrices\nare two classical combinatorial structures while one provides a more efficient\ndecoding algorithm and the other needs fewer tests, i.e., larger rate.\nRecently, a notion of strongly separable matrix has been introduced, which has\nthe same identifying ability as a disjunct matrix, but has larger rate. In this\npaper, we use a modified probabilistic method to improve the lower bounds for\nthe rate of strongly separable matrices. Using this method, we also improve the\nlower bounds for some well-known combinatorial structures, including locally\nthin set families and cancellative set families.\n", " A new generalization of the Gray map is introduced. The new generalization\n$\\Phi: Z_{2^k}^n \\to Z_{2}^{2^{k-1}n}$ is connected with the known generalized\nGray map $\\phi$ in the following way: if we take two dual linear\n$Z_{2^k}$-codes and construct binary codes from them using the generalizations\n$\\phi$ and $\\Phi$ of the Gray map, then the weight enumerators of the binary\ncodes obtained will satisfy the MacWilliams identity. The classes of\n$Z_{2^k}$-linear Hadamard codes and co-$Z_{2^k}$-linear extended 1-perfect\ncodes are described, where co-$Z_{2^k}$-linearity means that the code can be\nobtained from a linear $Z_{2^k}$-code with the help of the new generalized Gray\nmap. Keywords: Gray map, Hadamard codes, MacWilliams identity, perfect codes,\n$Z_{2^k}$-linearity\n", " We study a simplified, solvable model of a fully-connected metabolic network\nwith constrained quenched disorder to mimic the conservation laws imposed by\nstoichiometry on chemical reactions. Within a spin-glass type of approach, we\nshow that in presence of a conserved metabolic pool the flux state\ncorresponding to maximal growth is stationary independently of the pool size.\nIn addition, and at odds with the case of unconstrained networks, the volume of\noptimal flux configurations remains finite, indicating that the frustration\nimposed by stoichiometric constraints, while reducing growth capabilities,\nconfers robustness and flexibility to the system. These results have a clear\nbiological interpretation and provide a basic, fully analytical explanation to\nfeatures recently observed in real metabolic networks.\n", " We show that all the standard distances from metric geometry and functional\nanalysis, such as Gromov-Hausdorff distance, Banach-Mazur distance, Kadets\ndistance, Lipschitz distance, Net distance, and Hausdorff-Lipschitz distance\nhave all the same complexity and are reducible to each other in a precisely\ndefined way.\n This is done in terms of descriptive set theory and is a part of a larger\nresearch program initiated by the authors in \\emph{Complexity of distances:\nTheory of generalized analytic equivalence relations}. The paper is however\ntargeted also to specialists in metric geometry and geometry of Banach spaces.\n", " We construct bijections giving three \"codes\" for trees. These codes follow\nnaturally from the Matrix Tree Theorem of Tutte and have many advantages over\nthe one produced by Prufer in 1918. One algorithm gives explicitly a bijection\nthat is implicit in Orlin's manipulatorial proof of Cayley's formula (the\nformula was actually found first by Borchardt). Another is based on a proof of\nKnuth. The third is an implementation of Joyal's pseudo-bijective proof of the\nformula, and is equivalent to one previously found by Egecioglu and Remmel. In\neach case, we have at least two algorithms, one of which involves hands-on\nmanipulations of the tree while the other involves a combinatorial and linear\nalgebraic manipulation of a matrix.\n", " In recent years, new methods have been developed to synthesize complex porous\nand micro-structured geometry in a variety of ways. In this work, we take these\napproaches one step further and present these methods as an efficacious design\nparadigm. Specifically, complex micro-structure geometry can be synthesized\nwhile optimizing certain properties such as maximal heat exchange in heat\nexchangers, or minimal weight under stress specifications. By being able to\nadjust the geometry, the topology and/or the material properties of individual\ntiles in the micro-structure, possibly in a gradual way, a porous object can be\nsynthesized that is optimal with respect to the design specifications. As part\nof this work, we exemplify this paradigm on a variety of diverse applications.\n" ]

[ " In this article we present a class of codes with few weights arising from\nspecial type of linear sets. We explicitly show the weights of such codes,\ntheir weight enumerator and possible choices for their generator matrices. In\nparticular, our construction yields also to linear codes with three weights\nand, in some cases, to almost MDS codes. The interest for these codes relies on\ntheir applications to authentication codes and secret schemes, and their\nconnections with further objects such as association schemes and graphs.\n", " Over the past few decades, there has been extensive research on scattered\nsubspaces, partly because of their link to MRD codes. These subspaces can be\ncharacterized using linearized polynomials over finite fields. Within this\ncontext, scattered sequences extend the concept of scattered polynomials and\ncan be viewed as geometric equivalents of exceptional MRD codes. Up to now,\nonly scattered sequences of orders one and two have been developed. However,\nthis paper presents an infinite series of exceptional scattered sequences of\nany order beyond two which correspond to scattered subspaces that cannot be\nobtained as direct sum of scattered subspaces in smaller dimensions. The paper\nalso addresses equivalence concerns within this framework.\n", " Federated Learning (FL) has emerged as a promising technique for edge devices\nto collaboratively learn a shared machine learning model while keeping training\ndata locally on the device, thereby removing the need to store and access the\nfull data in the cloud. However, FL is difficult to implement, test and deploy\nin practice considering heterogeneity in common edge device settings, making it\nfundamentally hard for researchers to efficiently prototype and test their\noptimization algorithms. In this work, our aim is to alleviate this problem by\nintroducing FL_PyTorch : a suite of open-source software written in python that\nbuilds on top of one the most popular research Deep Learning (DL) framework\nPyTorch. We built FL_PyTorch as a research simulator for FL to enable fast\ndevelopment, prototyping and experimenting with new and existing FL\noptimization algorithms. Our system supports abstractions that provide\nresearchers with a sufficient level of flexibility to experiment with existing\nand novel approaches to advance the state-of-the-art. Furthermore, FL_PyTorch\nis a simple to use console system, allows to run several clients simultaneously\nusing local CPUs or GPU(s), and even remote compute devices without the need\nfor any distributed implementation provided by the user. FL_PyTorch also offers\na Graphical User Interface. For new methods, researchers only provide the\ncentralized implementation of their algorithm. To showcase the possibilities\nand usefulness of our system, we experiment with several well-known\nstate-of-the-art FL algorithms and a few of the most common FL datasets.\n", " We construct bijections giving three \"codes\" for trees. These codes follow\nnaturally from the Matrix Tree Theorem of Tutte and have many advantages over\nthe one produced by Prufer in 1918. One algorithm gives explicitly a bijection\nthat is implicit in Orlin's manipulatorial proof of Cayley's formula (the\nformula was actually found first by Borchardt). Another is based on a proof of\nKnuth. The third is an implementation of Joyal's pseudo-bijective proof of the\nformula, and is equivalent to one previously found by Egecioglu and Remmel. In\neach case, we have at least two algorithms, one of which involves hands-on\nmanipulations of the tree while the other involves a combinatorial and linear\nalgebraic manipulation of a matrix.\n", " Conserved moieties are groups of atoms that remain intact in all reactions of\na metabolic network. Identification of conserved moieties gives insight into\nthe structure and function of metabolic networks and facilitates metabolic\nmodelling. All moiety conservation relations can be represented as nonnegative\ninteger vectors in the left null space of the stoichiometric matrix\ncorresponding to a biochemical network. Algorithms exist to compute such\nvectors based only on reaction stoichiometry but their computational complexity\nhas limited their application to relatively small metabolic networks. Moreover,\nthe vectors returned by existing algorithms do not, in general, represent\nconservation of a specific moiety with a defined atomic structure. Here, we\nshow that identification of conserved moieties requires data on reaction atom\nmappings in addition to stoichiometry. We present a novel method to identify\nconserved moieties in metabolic networks by graph theoretical analysis of their\nunderlying atom transition networks. Our method returns the exact group of\natoms belonging to each conserved moiety as well as the corresponding vector in\nthe left null space of the stoichiometric matrix. It can be implemented as a\npipeline of polynomial time algorithms. Our implementation completes in under\nfive minutes on a metabolic network with more than 4,000 mass balanced\nreactions. The scalability of the method enables extension of existing\napplications for moiety conservation relations to genome-scale metabolic\nnetworks. We also give examples of new applications made possible by\nelucidating the atomic structure of conserved moieties.\n", " This paper introduces Whittemore, a language for causal programming. Causal\nprogramming is based on the theory of structural causal models and consists of\ntwo primary operations: identification, which finds formulas that compute\ncausal queries, and estimation, which applies formulas to transform probability\ndistributions to other probability distribution. Causal programming provides\nabstractions to declare models, queries, and distributions with syntax similar\nto standard mathematical notation, and conducts rigorous causal inference,\nwithout requiring detailed knowledge of the underlying algorithms. Examples of\ncausal inference with real data are provided, along with discussion of the\nimplementation and possibilities for future extension.\n", " We report some computational results comparing parallel and sequential codes\nfor vertex/facet enumeration problems for convex polyhedra. The problems chosen\nspan the range from simple to highly degenerate polytopes. We tested one code\n(lrs) based on pivoting and four codes (cddr+, ppl, normaliz, PORTA) based on\nthe double description method. normaliz employs parallelization as do the codes\nplrs and mplrs which are based on lrs. We tested these codes using various\nhardware configurations with up to 1200 cores. Major speedups were obtained by\nparallelization, particularly by the code mplrs which uses MPI and can operate\non clusters of machines.\n", " The study of control mechanisms of biological systems allows for interesting\napplications in bioengineering and medicine, for instance in cell reprogramming\nor drug target identification. A control strategy often consists of a set of\ninterventions that, by fixing the values of some components, ensure that the\nlong term dynamics of the controlled system is in a desired state. A common\napproach to control in the Boolean framework consists in checking how the fixed\nvalues propagate through the network, to establish whether the effect of\npercolating the interventions is sufficient to induce the target state.\nAlthough methods based uniquely on value percolation allow for efficient\ncomputation, they can miss many control strategies. Exhaustive methods for\ncontrol strategy identification, on the other hand, often entail high\ncomputational costs. In order to increase the number of control strategies\nidentified while still benefiting from an efficient implementation, we\nintroduce a method based on value percolation that uses trap spaces, subspaces\nof the state space that are closed with respect to the dynamics, and that can\nusually be easily computed in biological networks. The approach allows for node\ninterventions, which fix the value of certain components, and edge\ninterventions, which fix the effect that one component has on another. The\nmethod is implemented using Answer Set Programming, extending an existing\nefficient implementation of value percolation to allow for the use of trap\nspaces and edge control. The applicability of the approach is studied for\ndifferent control targets in a biological case study, identifying in all cases\nnew control strategies that would escape usual percolation-based methods.\n", " A vibrant theoretical research area are efficient exact parameterized\nalgorithms. Very recent solving competitions such as the PACE challenge show\nthat there is also increasing practical interest in the parameterized\nalgorithms community. An important research question is whether dedicated\nparameterized exact algorithms exhibit certain practical relevance and one can\neven beat well-established problem solvers. We consider the logic-based\ndeclarative modeling language and problem solving framework Answer Set\nProgramming (ASP). State-of-the-art ASP solvers rely considerably on Sat-based\nalgorithms. An ASP solver (DynASP2), which is based on a classical dynamic\nprogramming on tree decompositions, has been published very recently.\nUnfortunately, DynASP2 can outperform modern ASP solvers on programs of small\ntreewidth only if the question of interest is to count the number of solutions.\nIn this paper, we describe underlying concepts of our new implementation\n(DynASP2.5) that shows competitive behavior to state-of-the-art ASP solvers\neven for finding just one solution when solving problems as the Steiner tree\nproblem that have been modeled in ASP on graphs with low treewidth. Our\nimplementation is based on a novel approach that we call multi-pass dynamic\nprogramming (M-DPSINC).\n", " The identification of key factors such as medications, diseases, and\nrelationships within electronic health records and clinical notes has a wide\nrange of applications in the clinical field. In the N2C2 2022 competitions,\nvarious tasks were presented to promote the identification of key factors in\nelectronic health records (EHRs) using the Contextualized Medication Event\nDataset (CMED). Pretrained large language models (LLMs) demonstrated\nexceptional performance in these tasks. This study aims to explore the\nutilization of LLMs, specifically ChatGPT, for data augmentation to overcome\nthe limited availability of annotated data for identifying the key factors in\nEHRs. Additionally, different pre-trained BERT models, initially trained on\nextensive datasets like Wikipedia and MIMIC, were employed to develop models\nfor identifying these key variables in EHRs through fine-tuning on augmented\ndatasets. The experimental results of two EHR analysis tasks, namely medication\nidentification and medication event classification, indicate that data\naugmentation based on ChatGPT proves beneficial in improving performance for\nboth medication identification and medication event classification.\n", " In this paper, we define Abelian and consta-Abelian polyadic codes over rings\ndefined as affine algebras over chain rings. For that aim, we use the classical\nconstruction via splittings and multipliers of the underlying Abelian group. We\nalso derive some results on the structure of the associated polyadic codes and\nthe number of codes under these conditions.\n", " Background: The study of genome-scale metabolic models and their underlying\nnetworks is one of the most important fields in systems biology. The complexity\nof these models and their description makes the use of computational tools an\nessential element in their research. Therefore there is a strong need of\nefficient and versatile computational tools for the research in this area.\n Results: In this manuscript we present PyNetMet, a Python library of tools to\nwork with networks and metabolic models. These are open-source free tools for\nuse in a Python platform, which adds considerably versatility to them when\ncompared with their desktop software similars. On the other hand these tools\nallow one to work with different standards of metabolic models (OptGene and\nSBML) and the fact that they are programmed in Python opens the possibility of\nefficient integration with any other already existing Python tool.\n Conclusions: PyNetMet is, therefore, a collection of computational tools that\nwill facilitate the research work with metabolic models and networks.\n", " The automatic identification of harmful content online is of major concern\nfor social media platforms, policymakers, and society. Researchers have studied\ntextual, visual, and audio content, but typically in isolation. Yet, harmful\ncontent often combines multiple modalities, as in the case of memes, which are\nof particular interest due to their viral nature. With this in mind, here we\noffer a comprehensive survey with a focus on harmful memes. Based on a\nsystematic analysis of recent literature, we first propose a new typology of\nharmful memes, and then we highlight and summarize the relevant state of the\nart. One interesting finding is that many types of harmful memes are not really\nstudied, e.g., such featuring self-harm and extremism, partly due to the lack\nof suitable datasets. We further find that existing datasets mostly capture\nmulti-class scenarios, which are not inclusive of the affective spectrum that\nmemes can represent. Another observation is that memes can propagate globally\nthrough repackaging in different languages and that they can also be\nmultilingual, blending different cultures. We conclude by highlighting several\nchallenges related to multimodal semiotics, technological constraints, and\nnon-trivial social engagement, and we present several open-ended aspects such\nas delineating online harm and empirically examining related frameworks and\nassistive interventions, which we believe will motivate and drive future\nresearch.\n", " We present some open problems and describe briefly some possible research\ndirections in the emerging theory of Hardy spaces of Dirichlet series and their\nintimate counterparts, Hardy spaces on the infinite-dimensional torus. Links to\nnumber theory are emphasized throughout the paper.\n", " These two lectures cover some of the advances that underpin recent progress\nin deriving continuum solutions from the exact renormalization group. We\nconcentrate on concepts and on exact non-perturbative statements, but in the\nprocess will describe how real non-perturbative calculations can be done,\nparticularly within derivative expansion approximations. An effort has been\nmade to keep the lectures pedagogical and self-contained. Topics covered are\nthe derivation of the flow equations, their equivalence, continuum limits,\nperturbation theory, truncations, derivative expansions, identification of\nfixed points and eigenoperators, and the role of reparametrization invariance.\nSome new material is included, in particular a demonstration of\nnon-perturbative renormalizability, and a discussion of ultraviolet\nrenormalons.\n", " Neural codes serve as a language for neurons in the brain. Convex codes,\nwhich arise from the pattern of intersections of convex sets in Euclidean\nspace, are of particular relevance to neuroscience. Not every code is convex,\nhowever, and the combinatorial properties of a code that determine its\nconvexity are still poorly understood. Here we find that a code that can be\nrealized by a collection of open convex sets may or may not be realizable by\nclosed convex sets, and vice versa, establishing that open convex and closed\nconvex codes are distinct classes. We also prove that max intersection-complete\ncodes (i.e. codes that contain all intersections of maximal codewords) are both\nopen convex and closed convex, and provide an upper bound for their minimal\nembedding dimension. Finally, we show that the addition of non-maximal\ncodewords to an open convex code preserves convexity.\n", " Previous work on convexity of neural codes has produced codes that are\nopen-convex but not closed-convex -- or vice-versa. However, why a code is one\nbut not the other, and how to detect such discrepancies are open questions. We\ntackle these questions in two ways. First, we investigate the concept of\ndegeneracy introduced by Cruz et al., and extend their results to show that\nnondegeneracy precisely captures the situation when taking closures or\ninteriors of open or closed realizations, respectively, yields another\nrealization of the code. Second, we give the first general criteria for\nprecluding a code from being closed-convex (without ruling out open-convexity),\nunifying ad-hoc geometric arguments in prior works. One criterion is built on a\nphenomenon we call a rigid structure, while the other can be stated\nalgebraically, in terms of the neural ideal of the code. These results\ncomplement existing criteria having the opposite purpose: precluding\nopen-convexity but not closed-convexity. Finally, we show that a family of\ncodes shown by Jeffs to be not open-convex is in fact closed-convex and\nrealizable in dimension two.\n", " The exploration of linear subspaces, particularly scattered subspaces, has\ngarnered considerable attention across diverse mathematical disciplines in\nrecent years, notably within finite geometries and coding theory. Scattered\nsubspaces play a pivotal role in analyzing various geometric structures such as\nblocking sets, two-intersection sets, complete arcs, caps in affine and\nprojective spaces over finite fields and rank metric codes. This paper\nintroduces a new infinite family of $h$-subspaces, along with their associated\nMRD codes. Additionally, it addresses the task of determining the generalized\nweights of these codes. Notably, we demonstrate that these MRD codes exhibit\nsome larger generalized weights compared to those previously identified.\n", " We study circle valued maps and consider the persistence of the homology of\ntheir fibers. The outcome is a finite collection of computable invariants which\nanswer the basic questions on persistence and in addition encode the topology\nof the source space and its relevant subspaces. Unlike persistence of real\nvalued maps, circle valued maps enjoy a different class of invariants called\nJordan cells in addition to bar codes. We establish a relation between the\nhomology of the source space and of its relevant subspaces with these\ninvariants and provide a new algorithm to compute these invariants from an\ninput matrix that encodes a circle valued map on an input simplicial complex.\n", " Convolutional codes are constructed, designed and analysed using row and/or\nblock structures of unit algebraic schemes. Infinite series of such codes and\nof codes with specific properties are derived. Properties are shown\nalgebraically and algebraic decoding methods are derived. For a given rate and\ngiven error-correction capability at each component, convolutional codes with\nthese specifications and with efficient decoding algorithms are constructed.\nExplicit prototype examples are given but in general large lengths and large\nerror capability are achievable. Convolutional codes with efficient decoding\nalgorithms at or near the maximum free distances attainable for the parameters\nare constructible. Unit memory convolutional codes of maximum possible free\ndistance are designed with practical algebraic decoding algorithms.\n LDPC (low density parity check) convolutional codes with efficient decoding\nschemes are constructed and analysed by the methods. Self-dual and\ndual-containing convolutional codes may also be designed by the methods;\ndual-containing codes enables the construction of quantum codes.\n", " Explainable Artificial Intelligence (XAI) has re-emerged in response to the\ndevelopment of modern AI and ML systems. These systems are complex and\nsometimes biased, but they nevertheless make decisions that impact our lives.\nXAI systems are frequently algorithm-focused; starting and ending with an\nalgorithm that implements a basic untested idea about explainability. These\nsystems are often not tested to determine whether the algorithm helps users\naccomplish any goals, and so their explainability remains unproven. We propose\nan alternative: to start with human-focused principles for the design, testing,\nand implementation of XAI systems, and implement algorithms to serve that\npurpose. In this paper, we review some of the basic concepts that have been\nused for user-centered XAI systems over the past 40 years of research. Based on\nthese, we describe the \"Self-Explanation Scorecard\", which can help developers\nunderstand how they can empower users by enabling self-explanation. Finally, we\npresent a set of empirically-grounded, user-centered design principles that may\nguide developers to create successful explainable systems.\n", " Fast and reliable K-Nearest Neighbor Graph algorithms are more important than\never due to their widespread use in many data processing techniques. This paper\npresents a runtime optimized C implementation of the heuristic \"NN-Descent\"\nalgorithm by Wei Dong et al. for the l2-distance metric. Various implementation\noptimizations are explained which improve performance for low-dimensional as\nwell as high dimensional datasets. Optimizations to speed up the selection of\nwhich datapoint pairs to evaluate the distance for are primarily impactful for\nlow-dimensional datasets. A heuristic which exploits the iterative nature of\nNN-Descent to reorder data in memory is presented which enables better use of\nlocality and thereby improves the runtime. The restriction to the l2-distance\nmetric allows for the use of blocked distance evaluations which significantly\nincrease performance for high dimensional datasets. In combination the\noptimizations yield an implementation which significantly outperforms a widely\nused implementation of NN-Descent on all considered datasets. For instance, the\nruntime on the popular MNIST handwritten digits dataset is halved.\n", " Graph neural networks (GNNs) demonstrate great performance in compound\nproperty and activity prediction due to their capability to efficiently learn\ncomplex molecular graph structures. However, two main limitations persist\nincluding compound representation and model interpretability. While atom-level\nmolecular graph representations are commonly used because of their ability to\ncapture natural topology, they may not fully express important substructures or\nfunctional groups which significantly influence molecular properties.\nConsequently, recent research proposes alternative representations employing\nreduction techniques to integrate higher-level information and leverages both\nrepresentations for model learning. However, there is still a lack of study\nabout different molecular graph representations on model learning and\ninterpretation. Interpretability is also crucial for drug discovery as it can\noffer chemical insights and inspiration for optimization. Numerous studies\nattempt to include model interpretation to explain the rationale behind\npredictions, but most of them focus solely on individual prediction with little\nanalysis of the interpretation on different molecular graph representations.\nThis research introduces multiple molecular graph representations that\nincorporate higher-level information and investigates their effects on model\nlearning and interpretation from diverse perspectives. The results indicate\nthat combining atom graph representation with reduced molecular graph\nrepresentation can yield promising model performance. Furthermore, the\ninterpretation results can provide significant features and potential\nsubstructures consistently aligning with background knowledge. These multiple\nmolecular graph representations and interpretation analysis can bolster model\ncomprehension and facilitate relevant applications in drug discovery.\n", " As LLMs make their way into many aspects of our lives, one place that\nwarrants increased scrutiny with LLM usage is scientific research. Using LLMs\nfor generating or analyzing data for research purposes is gaining popularity.\nBut when such application is marred with ad-hoc decisions and engineering\nsolutions, we need to be concerned about how it may affect that research, its\nfindings, or any future works based on that research. We need a more scientific\napproach to using LLMs in our research. While there are several active efforts\nto support more systematic construction of prompts, they are often focused more\non achieving desirable outcomes rather than producing replicable and\ngeneralizable knowledge with sufficient transparency, objectivity, or rigor.\nThis article presents a new methodology inspired by codebook construction\nthrough qualitative methods to address that. Using humans in the loop and a\nmulti-phase verification processes, this methodology lays a foundation for more\nsystematic, objective, and trustworthy way of applying LLMs for analyzing data.\nSpecifically, we show how a set of researchers can work through a rigorous\nprocess of labeling, deliberating, and documenting to remove subjectivity and\nbring transparency and replicability to prompt generation process. A set of\nexperiments are presented to show how this methodology can be put in practice.\n", " Understanding control mechanisms in biological systems plays a crucial role\nin important applications, for instance in cell reprogramming. Boolean modeling\nallows the identification of possible efficient strategies, helping to reduce\nthe usually high and time-consuming experimental efforts. Available approaches\nto control strategy identification usually focus either on attractor or\nphenotype control, and are unable to deal with more complex control problems,\nfor instance phenotype avoidance. They also fail to capture, in many\nsituations, all possible minimal strategies, finding instead only sub-optimal\nsolutions. In order to fill these gaps, we present a novel approach to control\nstrategy identification in Boolean networks based on model checking. The method\nis guaranteed to identify all minimal control strategies, and provides maximal\nflexibility in the definition of the control target. We investigate the\napplicability of the approach by considering a range of control problems for\ndifferent biological systems, comparing the results, where possible, to those\nobtained by alternative control methods.\n", " In the present paper we introduce and study finite point subsets of a special\nkind, called optimum distributions, in the n-dimensional unit cube. Such\ndistributions are closely related with known (delta,s,n)-nets of low\ndiscrepancy. It turns out that optimum distributions have a rich combinatorial\nstructure. Namely, we show that optimum distributions can be characterized\ncompletely as maximum distance separable codes with respect to a non-Hamming\nmetric. Weight spectra of such codes can be evaluated precisely. We also\nconsider linear codes and distributions and study their general properties\nincluding the duality with respect to a suitable inner product. The\ncorresponding generalized MacWilliams identities for weight enumerators are\nbriefly discussed. Broad classes of linear maximum distance separable codes and\nlinear optimum distributions are explicitely constructed in the paper by the\nHermite interpolations over finite fields.\n", " Boundary labeling deals with annotating features in images such that labels\nare placed outside of the image and are connected by curves (so-called leaders)\nto the corresponding features. While boundary labeling has been extensively\ninvestigated from an algorithmic perspective, the research on its readability\nhas been neglected. In this paper we present the first formal user study on the\nreadability of boundary labeling. We consider the four most studied leader\ntypes with respect to their performance, i.e., whether and how fast a viewer\ncan assign a feature to its label and vice versa. We give a detailed analysis\nof the results regarding the readability of the four models and discuss their\naesthetic qualities based on the users' preference judgments and interviews.\n", " This paper presents a fast implementation of the Graham scan on the GPU. The\nproposed algorithm is composed of two stages: (1) two rounds of preprocessing\nperformed on the GPU and (2) the finalization of finding the convex hull on the\nCPU. We first discard the interior points that locate inside a quadrilateral\nformed by four extreme points, sort the remaining points according to the\nangles, and then divide them into the left and the right regions. For each\nregion, we perform a second round of filtering using the proposed preprocessing\napproach to discard the interior points in further. We finally obtain the\nexpected convex hull by calculating the convex hull of the remaining points on\nthe CPU. We directly employ the parallel sorting, reduction, and partitioning\nprovided by the library Thrust for better efficiency and simplicity.\nExperimental results show that our implementation achieves 6x ~ 7x speedups\nover the Qhull implementation for 20M points.\n", " We describe a coupled cluster framework for coupled systems of electrons and\nphonons. Neutral and charged excitations are accessed via the\nequation-of-motion version of the theory. Benchmarks on the Hubbard-Holstein\nmodel allow us to assess the strengths and weaknesses of different coupled\ncluster approximations which generally perform well for weak to moderate\ncoupling. Finally, we report progress towards an implementation for {\\it ab\ninitio} calculations on solids, and present some preliminary results on\nfinite-size models of diamond. We also report the implementation of\nelectron-phonon coupling matrix elements from crystalline Gaussian type\norbitals (cGTO) within the PySCF program package.\n", " Metabolite structure identification has become the major bottleneck of the\nmass spectrometry based metabolomics research. Till now, number of mass spectra\ndatabases and search algorithms have been developed to address this issue.\nHowever, two critical problems still exist: the low chemical component record\ncoverage in databases and significant MS/MS spectra variations related to\nexperiment equipment and parameter settings. In this work, we considered the\nmolecule fragment as basic building blocks of the metabolic components which\nhad relatively consistent signatures in MS/MS spectra. And from a bottom-up\npoint of view, we built a fragment centered database, MSFragDB, by reorganizing\nthe data from the Human Metabolome Database (HMDB) and developed an\nintensity-free searching algorithm to search and rank the most relative\nmetabolite according to the users' input. We also proposed the concept of\nfragment network, a graph structure that encoded the relationship between the\nmolecule fragments to find close motif that indicated a specific chemical\nstructure. Although based on the same dataset as the HMDB, validation results\nimplied that the MSFragDB had a higher hit ratio and furthermore, estimated\npossible taxonomy that a query spectrum belongs to when the corresponding\nchemical component was missing in the database. Aid by the Fragment Network,\nthe MSFragDB was also proved to be able to estimate the right structure while\nthe MS/MS spectrum suffers from the precursor-contamination. The strategy\nproposed is general and can be adopted in existing databases. We believe\nMSFragDB and Fragment Network can improve the performance of structure\nidentification with existing data. The beta version of the database is freely\navailable at www.xrzhanglab.com/msfragdb/.\n", " We examine the modular structure of the metabolic network when combined with\nthe regulatory network representing direct regulation of enzymes by small\nmetabolites in E.coli. In order to identify the modular structure we introduce\nclustering algorithm based on a novel vertex similarity measure for bipartite\ngraphs. We also apply a standard module identification method based on\nsimulated annealing. Both methods identify the same modular core each of them\nwith different resolution. We observe slight but still statistically\nsignificant increase of modularity after regulatory interactions addition.\nEnrichment of the metabolic network with the regulatory information leads to\nidentification of new functional modules, which cannot be detected in the\nmetabolic network only. Regulatory loops in the modules are the source of their\nself-control, i.e. autonomy, and allow to make hypothesis about module\nfunction. This study demonstrates that incorporation of regulatory information\nis important component in defining functional units of the metabolic network.\n", " The aim of this work is to give degree formulas for the generalized Hamming\nweights of evaluation codes and to show lower bounds for these weights. In\nparticular, we give degree formulas for the generalized Hamming weights of\nReed--Muller-type codes, and we determine the minimum distance of toric codes\nover hypersimplices, and the 1st and 2nd generalized Hamming weights of\nsquarefree evaluation codes.\n", " Vision-and-Language Navigation (VLN) aims to develop embodied agents that\nnavigate based on human instructions. However, current VLN frameworks often\nrely on static environments and optimal expert supervision, limiting their\nreal-world applicability. To address this, we introduce Human-Aware\nVision-and-Language Navigation (HA-VLN), extending traditional VLN by\nincorporating dynamic human activities and relaxing key assumptions. We propose\nthe Human-Aware 3D (HA3D) simulator, which combines dynamic human activities\nwith the Matterport3D dataset, and the Human-Aware Room-to-Room (HA-R2R)\ndataset, extending R2R with human activity descriptions. To tackle HA-VLN\nchallenges, we present the Expert-Supervised Cross-Modal (VLN-CM) and\nNon-Expert-Supervised Decision Transformer (VLN-DT) agents, utilizing\ncross-modal fusion and diverse training strategies for effective navigation in\ndynamic human environments. A comprehensive evaluation, including metrics\nconsidering human activities, and systematic analysis of HA-VLN's unique\nchallenges, underscores the need for further research to enhance HA-VLN agents'\nreal-world robustness and adaptability. Ultimately, this work provides\nbenchmarks and insights for future research on embodied AI and Sim2Real\ntransfer, paving the way for more realistic and applicable VLN systems in\nhuman-populated environments.\n", " In this paper we characterize the algebraic structure of two-dimensional\n$(\\alpha,\\beta )$-constacyclic codes of arbitrary length $s.\\ell$ and of their\nduals. For $\\alpha,\\beta \\in \\{1,-1\\}$, we give necessary and sufficient\nconditions for a two-dimensional $(\\alpha,\\beta )$-constacyclic code to be\nself-dual. We also show that a two-dimensional $(\\alpha,1 )$-constacyclic code\n$\\mathcal{C}$ of length $n=s.\\ell$ can not be self-dual if $\\gcd(s,q)= 1$.\nFinally, we give some examples of self-dual, isodual, MDS and quasi-twisted\ncodes corresponding to two-dimensional $(\\alpha,\\beta )$-constacyclic codes.\n", " Knowledge graphs (KGs) store highly heterogeneous information about the world\nin the structure of a graph, and are useful for tasks such as question\nanswering and reasoning. However, they often contain errors and are missing\ninformation. Vibrant research in KG refinement has worked to resolve these\nissues, tailoring techniques to either detect specific types of errors or\ncomplete a KG.\n In this work, we introduce a unified solution to KG characterization by\nformulating the problem as unsupervised KG summarization with a set of\ninductive, soft rules, which describe what is normal in a KG, and thus can be\nused to identify what is abnormal, whether it be strange or missing. Unlike\nfirst-order logic rules, our rules are labeled, rooted graphs, i.e., patterns\nthat describe the expected neighborhood around a (seen or unseen) node, based\non its type, and information in the KG. Stepping away from the traditional\nsupport/confidence-based rule mining techniques, we propose KGist, Knowledge\nGraph Inductive SummarizaTion, which learns a summary of inductive rules that\nbest compress the KG according to the Minimum Description Length principle---a\nformulation that we are the first to use in the context of KG rule mining. We\napply our rules to three large KGs (NELL, DBpedia, and Yago), and tasks such as\ncompression, various types of error detection, and identification of incomplete\ninformation. We show that KGist outperforms task-specific, supervised and\nunsupervised baselines in error detection and incompleteness identification,\n(identifying the location of up to 93% of missing entities---over 10% more than\nbaselines), while also being efficient for large knowledge graphs.\n", " A new approach for the approximation of the channel log-likelihood ratio\n(LLR) for wireless channels based on Taylor series is proposed. The\napproximation is applied to the uncorrelated flat Rayleigh fading channel with\nunknown channel state information at the receiver. It is shown that the\nproposed approximation greatly simplifies the calculation of channel LLRs, and\nyet provides results almost identical to those based on the exact calculation\nof channel LLRs. The results are obtained in the context of bit-interleaved\ncoded modulation (BICM) schemes with low-density parity-check (LDPC) codes, and\ninclude threshold calculations and error rate performance of finite-length\ncodes. Compared to the existing approximations, the proposed method is either\nsignificantly less complex, or considerably more accurate.\n", " In the present paper, we propose a novel generalization of the celebrated MMP\nalgorithm in order to find the wavefront propagation and the cut-locus on a\nconvex polyhedron with an emphasis on actual implementation for instantaneous\nvisualization and numerical computation.\n", " We review some of the features of the evolutions equations for transverse\nmomentum dependent parton distributions recently proposed by us. We briefly\ndescribe the new ingredients entering the equations and their relationship with\nordinary evolution equations. We comment on possible choices for the initial\nconditions and then show results for the evolved distributions obtained by\nnumerical implementation of the equations. By computing the average transverse\nmomentum at different scale we highlight some general properties of the\nsolutions.\n", " Networks of neurons in the brain encode preferred patterns of neural activity\nvia their synaptic connections. Despite receiving considerable attention, the\nprecise relationship between network connectivity and encoded patterns is still\npoorly understood. Here we consider this problem for networks of\nthreshold-linear neurons whose computational function is to learn and store a\nset of binary patterns (e.g., a neural code) as \"permitted sets\" of the\nnetwork. We introduce a simple Encoding Rule that selectively turns \"on\"\nsynapses between neurons that co-appear in one or more patterns. The rule uses\nsynapses that are binary, in the sense of having only two states (\"on\" or\n\"off\"), but also heterogeneous, with weights drawn from an underlying synaptic\nstrength matrix S. Our main results precisely describe the stored patterns that\nresult from the Encoding Rule -- including unintended \"spurious\" states -- and\ngive an explicit characterization of the dependence on S. In particular, we\nfind that binary patterns are successfully stored in these networks when the\nexcitatory connections between neurons are geometrically balanced -- i.e., they\nsatisfy a set of geometric constraints. Furthermore, we find that certain types\nof neural codes are \"natural\" in the context of these networks, meaning that\nthe full code can be accurately learned from a highly undersampled set of\npatterns. Interestingly, many commonly observed neural codes in cortical and\nhippocampal areas are natural in this sense. As an application, we construct\nnetworks that encode hippocampal place field codes nearly exactly, following\npresentation of only a small fraction of patterns. To obtain our results, we\nprove new theorems using classical ideas from convex and distance geometry,\nsuch as Cayley-Menger determinants, revealing a novel connection between these\nareas of mathematics and coding properties of neural networks.\n", " A set $C$ of unit vectors in $\\mathbb{R}^d$ is called an $L$-spherical code\nif $x \\cdot y \\in L$ for any distinct $x,y$ in $C$. Spherical codes have been\nextensively studied since their introduction in the 1970's by Delsarte,\nGoethals and Seidel. In this note we prove a conjecture of Bukh on the maximum\nsize of spherical codes. In particular, we show that for any set of $k$ fixed\nangles, one can choose at most $O(d^k)$ lines in $\\mathbb{R}^d$ such that any\npair of them forms one of these angles.\n", " The utilization of web mapping becomes increasingly important in the domain\nof cartography. Users want access to spatial data on the web specific to their\nneeds. For this reason, different approaches were appeared for generating\non-the-fly the maps demanded by users, but those not suffice for guide a\nflexible and efficient process. Thus, new approach must be developed for\nimproving this process according to the user needs. This work focuses on\ndefining a new strategy which improves on-the-fly map generalization process\nand resolves the spatial conflicts. This approach uses the multiple\nrepresentation and cartographic generalization. The map generalization process\nis based on the implementation of multi- agent system where each agent was\nequipped with a genetic patrimony.\n", " This study investigates the relationship between tweet sentiment across\ndiverse categories: news, company opinions, CEO opinions, competitor opinions,\nand stock market behavior in the biotechnology sector, with a focus on\nunderstanding the impact of social media discourse on investor sentiment and\ndecision-making processes. We analyzed historical stock market data for ten of\nthe largest and most influential pharmaceutical companies alongside Twitter\ndata related to COVID-19, vaccines, the companies, and their respective CEOs.\nUsing VADER sentiment analysis, we examined the sentiment scores of tweets and\nassessed their relationships with stock market performance. We employed ARIMA\n(AutoRegressive Integrated Moving Average) and VAR (Vector AutoRegression)\nmodels to forecast stock market performance, incorporating sentiment covariates\nto improve predictions. Our findings revealed a complex interplay between tweet\nsentiment, news, biotech companies, their CEOs, and stock market performance,\nemphasizing the importance of considering diverse factors when modeling and\npredicting stock prices. This study provides valuable insights into the\ninfluence of social media on the financial sector and lays a foundation for\nfuture research aimed at refining stock price prediction models.\n", " Intelligent Mesh Generation (IMG) represents a novel and promising field of\nresearch, utilizing machine learning techniques to generate meshes. Despite its\nrelative infancy, IMG has significantly broadened the adaptability and\npracticality of mesh generation techniques, delivering numerous breakthroughs\nand unveiling potential future pathways. However, a noticeable void exists in\nthe contemporary literature concerning comprehensive surveys of IMG methods.\nThis paper endeavors to fill this gap by providing a systematic and thorough\nsurvey of the current IMG landscape. With a focus on 113 preliminary IMG\nmethods, we undertake a meticulous analysis from various angles, encompassing\ncore algorithm techniques and their application scope, agent learning\nobjectives, data types, targeted challenges, as well as advantages and\nlimitations. We have curated and categorized the literature, proposing three\nunique taxonomies based on key techniques, output mesh unit elements, and\nrelevant input data types. This paper also underscores several promising future\nresearch directions and challenges in IMG. To augment reader accessibility, a\ndedicated IMG project page is available at\n\\url{https://github.com/xzb030/IMG_Survey}.\n", " One of the main problems of the research area of network coding is to compute\ngood lower and upper bounds of the achievable cardinality of so-called subspace\ncodes in $\\operatorname{PG}(n,q)$, i.e., the set of subspaces of\n$\\mathbb{F}_q^n$, for a given minimal distance. Here we generalize a\nconstruction of Etzion and Silberstein to a wide range of parameters. This\nconstruction, named coset construction, improves or attains several of the\npreviously best-known subspace code sizes and attains the MRD bound for an\ninfinite family of parameters.\n", " The range, segment and rectangle query problems are fundamental problems in\ncomputational geometry, and have extensive applications in many domains.\nDespite the significant theoretical work on these problems, efficient\nimplementations can be complicated. We know of very few practical\nimplementations of the algorithms in parallel, and most implementations do not\nhave tight theoretical bounds. We focus on simple and efficient parallel\nalgorithms and implementations for these queries, which have tight worst-case\nbound in theory and good parallel performance in practice. We propose to use a\nsimple framework (the augmented map) to model the problem. Based on the\naugmented map interface, we develop both multi-level tree structures and\nsweepline algorithms supporting range, segment and rectangle queries in two\ndimensions. For the sweepline algorithms, we propose a parallel paradigm and\nshow corresponding cost bounds. All of our data structures are work-efficient\nto build in theory and achieve a low parallel depth. The query time is almost\nlinear to the output size.\n We have implemented all the data structures described in the paper using a\nparallel augmented map library. Based on the library each data structure only\nrequires about 100 lines of C++ code. We test their performance on large data\nsets (up to $10^8$ elements) and a machine with 72-cores (144 hyperthreads).\nThe parallel construction achieves 32-68x speedup. Speedup numbers on queries\nare up to 126-fold. Our sequential implementation outperforms the CGAL library\nby at least 2x in both construction and queries. Our sequential implementation\ncan be slightly slower than the R-tree in the Boost library in some cases\n(0.6-2.5x), but has significantly better query performance (1.6-1400x) than\nBoost.\n", " In recent years, several algorithms for system identification with neural\nstate-space models have been introduced. Most of the proposed approaches are\naimed at reducing the computational complexity of the learning problem, by\nsplitting the optimization over short sub-sequences extracted from a longer\ntraining dataset. Different sequences are then processed simultaneously within\na minibatch, taking advantage of modern parallel hardware for deep learning. An\nissue arising in these methods is the need to assign an initial state for each\nof the sub-sequences, which is required to run simulations and thus to evaluate\nthe fitting loss. In this paper, we provide insights for calibration of neural\nstate-space training algorithms based on extensive experimentation and analyses\nperformed on two recognized system identification benchmarks. Particular focus\nis given to the choice and the role of the initial state estimation. We\ndemonstrate that advanced initial state estimation techniques are really\nrequired to achieve high performance on certain classes of dynamical systems,\nwhile for asymptotically stable ones basic procedures such as zero or random\ninitialization already yield competitive performance.\n", " Huffman Codes are optimal Instantaneous Fixed-to-Variable (FV) codes in which\nevery source symbol can only be encoded by one codeword. Relaxing these\nconstraints permits constructing better FV codes. More specifically, recent\nwork has shown that AIFV-$m$ codes can beat Huffman coding. AIFV-$m$ codes\nconstruct am $m$-tuple of different coding trees between which the code\nalternates and are only almost instantaneous (AI). This means that decoding a\nword might require a delay of a finite number of bits.\n Current algorithms for constructing optimal AIFV-$m$ codes are iterative\nprocesses that construct progressively \"better sets\" of code trees. The\nprocesses have been proven to finitely converge to the optimal code but with no\nknown bounds on the convergence rate.\n This paper derives a geometric interpretation of the space of binary AIFV-$2$\ncodes, permitting the development of the first polynomially time-bounded\niterative procedures for constructing optimal AIFV codes.\n We first show that a simple binary search procedure can replace the current\niterative process to construct optimal binary AIFV-$2$ codes. We then describe\nhow to frame the problem as a linear programming with an exponential number of\nconstraints but a polynomial-time separability oracle. This permits using the\nGr\\\"otschel, Lov\\'asz and Schrijver ellipsoid method to solve the problem in a\npolynomial number of steps. While more complicated, this second method has the\npotential to lead to a polynomial time algorithm to construct optimal AIFV-$m$\ncodes for general $m$.\n", " We study questions in incidence geometry where the precise position of points\nis `blurry' (e.g. due to noise, inaccuracy or error). Thus lines are replaced\nby narrow tubes, and more generally affine subspaces are replaced by their\nsmall neighborhood. We show that the presence of a sufficiently large number of\napproximately collinear triples in a set of points in d dimensional complex\nspace implies that the points are close to a low dimensional affine subspace.\nThis can be viewed as a stable variant of the Sylvester-Gallai theorem and its\nextensions.\n Building on the recently found connection between Sylvester-Gallai type\ntheorems and complex Locally Correctable Codes (LCCs), we define the new notion\nof stable LCCs, in which the (local) correction procedure can also handle small\nperturbations in the euclidean metric. We prove that such stable codes with\nconstant query complexity do not exist. No impossibility results were known in\nany such local setting for more than 2 queries.\n", " It is widely known in the machine learning community that class noise can be\n(and often is) detrimental to inducing a model of the data. Many current\napproaches use a single, often biased, measurement to determine if an instance\nis noisy. A biased measure may work well on certain data sets, but it can also\nbe less effective on a broader set of data sets. In this paper, we present\nnoise identification using classifier diversity (NICD) -- a method for deriving\na less biased noise measurement and integrating it into the learning process.\nTo lessen the bias of the noise measure, NICD selects a diverse set of\nclassifiers (based on their predictions of novel instances) to determine which\ninstances are noisy. We examine NICD as a technique for filtering, instance\nweighting, and selecting the base classifiers of a voting ensemble. We compare\nNICD with several other noise handling techniques that do not consider\nclassifier diversity on a set of 54 data sets and 5 learning algorithms. NICD\nsignificantly increases the classification accuracy over the other considered\napproaches and is effective across a broad set of data sets and learning\nalgorithms.\n", " In this paper we apply a specific type ANNs - convolutional neural networks\n(CNNs) - to the problem of finding start and endpoints of trends, which are the\noptimal points for entering and leaving the market. We aim to explore long-term\ntrends, which last several months, not days. The key distinction of our model\nis that its labels are fully based on expert opinion data. Despite the various\nmodels based solely on stock price data, some market experts still argue that\ntraders are able to see hidden opportunities. The labelling was done via the\nGUI interface, which means that the experts worked directly with images, not\nnumerical data. This fact makes CNN the natural choice of algorithm. The\nproposed framework requires the sequential interaction of three CNN submodels,\nwhich identify the presence of a changepoint in a window, locate it and finally\nrecognize the type of new tendency - upward, downward or flat. These submodels\nhave certain pitfalls, therefore the calibration of their hyperparameters is\nthe main direction of further research. The research addresses such issues as\nimbalanced datasets and contradicting labels, as well as the need for specific\nquality metrics to keep up with practical applicability. This paper is the full\ntext of the research, presented at the 20th International Conference on\nArtificial Intelligence and Soft Computing Web System (ICAISC 2021)\n" ]

[ " In this paper we develop a general theory of compressed sensing for analog\nsignals, in close similarity to prior results for vectors in finite dimensional\nspaces that are sparse in a given orthonormal basis. The signals are modeled by\nfunctions in a reproducing kernel Hilbert space. Sparsity is defined as the\nminimal number of terms in expansions based on the kernel functions. Minimizing\nthis number is under certain conditions equivalent to minimizing an atomic\nnorm, the pre-dual of the supremum norm for functions in the Hilbert space. The\nnorm minimizer is shown to exist based on a compactness argument. Recovery\nbased on minimizing the atomic norm is robust and stable, so it provides\ncontrollable accuracy for recovery when the signal is only approximately sparse\nand the measurement is corrupted by noise.\n As applications of the theory, we include results on the recovery of sparse\nbandlimited functions and functions that have a sparse inverse short-time\nFourier transform.\n", " Simulations of nuclear magnetic resonance (NMR) experiments can be an\nimportant tool for extracting information about molecular structure and\noptimizing experimental protocols but are often intractable on classical\ncomputers for large molecules such as proteins and for protocols such as\nzero-field NMR. We demonstrate the first quantum simulation of an NMR spectrum,\ncomputing the zero-field spectrum of the methyl group of acetonitrile using\nfour qubits of a trapped-ion quantum computer. We reduce the sampling cost of\nthe quantum simulation by an order of magnitude using compressed sensing\ntechniques. We show how the intrinsic decoherence of NMR systems may enable the\nzero-field simulation of classically hard molecules on relatively near-term\nquantum hardware and discuss how the experimentally demonstrated quantum\nalgorithm can be used to efficiently simulate scientifically and\ntechnologically relevant solid-state NMR experiments on more mature devices.\nOur work opens a practical application for quantum computation.\n", " Different space-filling curves (SFCs) are briefly reviewed in this paper, and\na new one is proposed. A century has passed between the inception of this kind\nof curves, since then they have been found useful in computer science,\nparticularly in data storage and indexing due to their clustering properties,\nbeing Hilbert curve the most well-known member of the family of fractals. The\nproposed Aztec curve, with similar characteristics to the Hilbert's curve, is\nintroduced in this paper, accompanied by a grammatical description for its\nconstruction. It yields the possibility of creating bi-dimensional clusters,\nnot available for Hilbert nor Peano curves. Additional to this, a case of\napplication on the scope of Compressed Sensing is implemented, in which the use\nof Hilbert curve is contrasted with Aztec curve, having a similar performance,\nand positioning the Aztec curve as viable and a new alternative for future\nexploitation on applications that make use of SFC's.\n", " We introduce the dune-curvilineargrid module. The module provides the\nself-contained, parallel grid manager, as well as the underlying elementary\ncurvilinear geometry module dune-curvilineargeometry. This work is motivated by\nthe need for reliable and scalable electromagnetic design of nanooptical\ndevices. Curvilinear geometries improve both the accuracy of modeling smooth\nmaterial boundaries, and the h/p-convergence rate of PDE solutions, reducing\nthe necessary computational effort. dune-curvilineargrid provides a large\nspectrum of features for scalable parallel implementations of Finite Element\nand Boundary Integral methods over curvilinear tetrahedral geometries,\nincluding symbolic polynomial mappings and operations, recursive integration,\nsparse and dense grid communication, parallel timing and memory footprint\ndiagnostics utilities. It is written in templated C++ using MPI for\nparallelization and ParMETIS for grid partitioning, and is provided as a module\nfor the DUNE interface. The dune-curvilineargrid grid manager is continuously\ndeveloped and improved, and so is this documentation. For the most recent\nversion of the documentation, as well as the source code, please refer to the\nprovided repositories and our website.\n", " The ability to perform sensitive, real-time, in situ, multiplex chemical\nanalysis is indispensable for diverse applications such as human health\nmonitoring, food safety testing, forensic analysis, environmental sensing, and\nhomeland security. Surface-enhanced Raman spectroscopy (SERS) is an effective\ntool to offer the ability by virtue of its high sensitivity and rapid\nlabel-free signal detection as well as the availability of portable Raman\nspectrometers. Unfortunately, the practical utility of SERS is limited because\nit generally requires sample collection and preparation, namely, collecting a\nsample from an object of interest and placing the sample on top of a SERS\nsubstrate to perform a SERS measurement. In fact, not all analytes can satisfy\nthis requirement because the sample collection and preparation process may be\nundesirable, laborious, difficult, dangerous, costly, or time-consuming. Here\nwe introduce \"Place & Play SERS\" based on an ultrathin, flexible, stretchable,\nadhesive, biointegratable gold-deposited polyvinyl alcohol (PVA) nanomesh\nsubstrate that enables placing the substrate on top of an object of interest\nand performing a SERS measurement of the object by epi-excitation without the\nneed for touching, destroying, and sampling it. Specifically, we characterized\nthe sensitivity of the gold/PVA nanomesh substrate in the Place & Play SERS\nmeasurement scheme and then used the scheme to conduct SERS measurements of\nboth wet and dry objects under nearly real-world conditions. To show the\npractical utility of Place & Play SERS, we demonstrated two examples of its\napplication: food safety testing and forensic analysis. Our results firmly\nverified the new measurement scheme of SERS and are expected to extend the\npotential of SERS by opening up untapped applications of sensitive, real-time,\nin situ multiplex chemical analysis.\n", " The time-domain cross-correlation of incoherent and random noise recorded by\na series of passive sensors contains the impulse response of the medium between\nthese sensors. By using noise generated by a can of compressed air sprayed on\nthe surface of a plexiglass plate, we are able to reconstruct not only the time\nof flight but the whole waveforms between the sensors. From the reconstruction\nof the direct $A_0$ and $S_0$ waves, we derive the dispersion curves of the\nflexural waves, thus estimating the mechanical properties of the material\nwithout a conventional electromechanical source. The dense array of receivers\nemployed here allow a precise frequency-wavenumber study of flexural waves,\nalong with a thorough evaluation of the rate of convergence of the correlation\nwith respect to the record length, the frequency, and the distance between the\nreceivers. The reconstruction of the actual amplitude and attenuation of the\nimpulse response is also addressed in this paper.\n", " This paper explores the genotype-phenotype relationship. It outlines\nconditions under which the dependence of a quantitative trait on the genome\nmight be predictable, based on measurement of a limited subset of genotypes. It\nuses the theory of real-valued Boolean functions in a systematic way to\ntranslate trait data into the Fourier domain. Important trait features, such as\nthe roughness of the trait landscape or the modularity of a trait have a simple\nFourier interpretation. Ruggedness at a gene location corresponds to high\nsensitivity to mutation, while a modular organization of gene activity reduces\nsuch sensitivity.\n Traits where rugged loci are rare will naturally compress gene data in the\nFourier domain, leading to a sparse representation of trait data, concentrated\nin identifiable, low-level coefficients. This Fourier representation of a trait\norganizes epistasis in a form which is isometric to the trait data. As Fourier\nmatrices are known to be maximally incoherent with the standard basis, this\npermits employing compressive sensing techniques to work from data sets that\nare relatively small -- sometimes even of polynomial size -- compared to the\nexponentially large sets of possible genomes.\n This theory provides a theoretical underpinning for systematic use of Boolean\nfunction machinery to dissect the dependency of a trait on the genome and\nenvironment.\n", " We combine the most recent observations of large-scale structure (2dF and\nSDSS galaxy surveys) and cosmic microwave anisotropies (WMAP and ACBAR) to put\nconstraints on flat cosmological models where the number of massive neutrinos\nand of massless relativistic relics are both left arbitrary. We discuss the\nimpact of each dataset and of various priors on our bounds. For the standard\ncase of three thermalized neutrinos, we find an upper bound on the total\nneutrino mass sum m_nu < 1.0 (resp. 0.6) eV (at 2sigma), using only CMB and LSS\ndata (resp. including priors from supernovae data and the HST Key Project), a\nbound that is quite insensitive to the splitting of the total mass between the\nthree species. When the total number of neutrinos or relativistic relics N_eff\nis left free, the upper bound on sum m_nu (at 2sigma, including all priors)\nranges from 1.0 to 1.5 eV depending on the mass splitting. We provide an\nexplanation of the parameter degeneracy that allows larger values of the masses\nwhen N_eff increases. Finally, we show that the limit on the total neutrino\nmass is not significantly modified in the presence of primordial gravitational\nwaves, because current data provide a clear distinction between the\ncorresponding effects.\n", " We investigate the computational complexity of tensor rank, a concept that\nplays fundamental role in different topics of modern applied mathematics. For\ntensors over any integral domain, we prove that the rank problem is polynomial\ntime equivalent to solving a system of polynomial equations over this integral\ndomain. Our result gives a complete description of the algorithmic complexity\nof tensor rank and allows one to solve several known open problems. In\nparticular, the tensor rank over $\\mathbb{Z}$ turns out to be undecidable,\nwhich answers the question posed by Gonzalez and Ja'Ja' in 1980. We generalize\nour result and prove that the symmetric rank admits a similar description of\ncomputational complexity as the one we give for usual rank. In particular,\ncomputing the symmetric rank of a rational tensor is shown to be NP-hard, which\nproves a recent conjecture of Hillar and Lim. As a byproduct of our approach,\nwe get a similar characterization of the algorithmic complexity of the minimal\nrank matrix completion problem, which gives a complete answer to the question\ndiscussed in 1999 by Buss, Frandsen, and Shallit.\n", " In this paper, we study the application of sparse principal component\nanalysis (PCA) to clustering and feature selection problems. Sparse PCA seeks\nsparse factors, or linear combinations of the data variables, explaining a\nmaximum amount of variance in the data while having only a limited number of\nnonzero coefficients. PCA is often used as a simple clustering technique and\nsparse factors allow us here to interpret the clusters in terms of a reduced\nset of variables. We begin with a brief introduction and motivation on sparse\nPCA and detail our implementation of the algorithm in d'Aspremont et al.\n(2005). We then apply these results to some classic clustering and feature\nselection problems arising in biology.\n", " Phylogenetic networks represent evolutionary history of species and can\nrecord natural reticulate evolutionary processes such as horizontal gene\ntransfer and gene recombination. This makes phylogenetic networks a more\ncomprehensive representation of evolutionary history compared to phylogenetic\ntrees. Stochastic processes for generating random trees or networks are\nimportant tools in evolutionary analysis, especially in phylogeny\nreconstruction where they can be utilized for validation or serve as priors for\nBayesian methods. However, as more network generators are developed, there is a\nlack of discussion or comparison for different generators. To bridge this gap,\nwe compare a set of phylogenetic network generators by profiling topological\nsummary statistics of the generated networks over the number of reticulations\nand comparing the topological profiles.\n", " The {\\em compressed stack} is a data structure designed by Barba {\\em et al.}\n(Algorithmica 2015) that allows to reduce the amount of memory needed by an\nalgorithm (at the cost of increasing its runtime). In this paper we introduce\nthe first implementation of this data structure and make its source code\npublicly available.\n Together with the implementation we analyze the performance of the compressed\nstack. In our synthetic experiments, considering different test scenarios and\nusing data sizes ranging up to $2^{30}$ elements, we compare it with the\nclassic (uncompressed) stack, both in terms of runtime and memory used.\n Our experiments show that the compressed stack needs significantly less\nmemory than the usual stack (this difference is significant for inputs\ncontaining $2000$ or more elements). Overall, with a proper choice of\nparameters, we can save a significant amount of space (from two to four orders\nof magnitude) with a small increase in the runtime ($2.32$ times slower on\naverage than the classic stack). These results holds even in test scenarios\nspecifically designed to be challenging for the compressed stack.\n", " For a large class of random matrices $A$ with i.i.d. entries we show that the\n$\\ell_1$-quotient property holds with probability exponentially close to 1. In\ncontrast to previous results, our analysis does not require concentration of\nthe entrywise distributions. We provide a unified proof that recovers\ncorresponding previous results for (sub-)Gaussian and Weibull distributions.\nOur findings generalize known results on the geometry of random polytopes,\nproviding lower bounds on the size of the largest Euclidean ball contained in\nthe centrally symmetric polytope spanned by the columns of $A$. At the same\ntime, our results establish robustness of noise-blind $\\ell_1$-decoders for\nrecovering sparse vectors $x$ from underdetermined, noisy linear measurements\n$y = Ax + w$ under the weakest possible assumptions on the entrywise\ndistributions that allow for recovery with optimal sample complexity even in\nthe noiseless case. Our analysis predicts superior robustness behavior for\nmeasurement matrices with super-Gaussian entries, which we confirm by numerical\nexperiments.\n", " We present the first sublinear memory sketch that can be queried to find the\nnearest neighbors in a dataset. Our online sketching algorithm compresses an N\nelement dataset to a sketch of size $O(N^b \\log^3 N)$ in $O(N^{(b+1)} \\log^3\nN)$ time, where $b < 1$. This sketch can correctly report the nearest neighbors\nof any query that satisfies a stability condition parameterized by $b$. We\nachieve sublinear memory performance on stable queries by combining recent\nadvances in locality sensitive hash (LSH)-based estimators, online kernel\ndensity estimation, and compressed sensing. Our theoretical results shed new\nlight on the memory-accuracy tradeoff for nearest neighbor search, and our\nsketch, which consists entirely of short integer arrays, has a variety of\nattractive features in practice. We evaluate the memory-recall tradeoff of our\nmethod on a friend recommendation task in the Google Plus social media network.\nWe obtain orders of magnitude better compression than the random projection\nbased alternative while retaining the ability to report the nearest neighbors\nof practical queries.\n", " There has been recently a growing interest for implicit shape\nrepresentations. Contrary to explicit representations, they have no resolution\nlimitations and they easily deal with a wide variety of surface topologies. To\nlearn these implicit representations, current approaches rely on a certain\nlevel of shape supervision (e.g., inside/outside information or\ndistance-to-shape knowledge), or at least require a dense point cloud (to\napproximate well enough the distance-to-shape). In contrast, we introduce\nNeeDrop, a self-supervised method for learning shape representations from\npossibly extremely sparse point clouds. Like in Buffon's needle problem, we\n\"drop\" (sample) needles on the point cloud and consider that, statistically,\nclose to the surface, the needle end points lie on opposite sides of the\nsurface. No shape knowledge is required and the point cloud can be highly\nsparse, e.g., as lidar point clouds acquired by vehicles. Previous\nself-supervised shape representation approaches fail to produce good-quality\nresults on this kind of data. We obtain quantitative results on par with\nexisting supervised approaches on shape reconstruction datasets and show\npromising qualitative results on hard autonomous driving datasets such as\nKITTI.\n", " Mixing (or quasirandom) properties of the natural transition matrix\nassociated to a graph can be quantified by its distance to the complete graph.\nDifferent mixing properties correspond to different norms to measure this\ndistance. For dense graphs, two such properties known as spectral expansion and\nuniformity were shown to be equivalent in seminal 1989 work of Chung, Graham\nand Wilson. Recently, Conlon and Zhao extended this equivalence to the case of\nsparse vertex transitive graphs using the famous Grothendieck inequality. Here\nwe generalize these results to the non-commutative, or `quantum', case, where a\ntransition matrix becomes a quantum channel. In particular, we show that for\nirreducibly covariant quantum channels, expansion is equivalent to a natural\nanalog of uniformity for graphs, generalizing the result of Conlon and Zhao.\nMoreover, we show that in these results, the non-commutative and commutative\n(resp.) Grothendieck inequalities yield the best-possible constants.\n", " We design a homotopy continuation algorithm, that is based on numerically\ntracking Viro's patchworking method, for finding real zeros of sparse\npolynomial systems. The algorithm is targeted for polynomial systems with\ncoefficients satisfying certain concavity conditions. It operates entirely over\nthe real numbers and tracks the optimal number of solution paths. In more\ntechnical terms; we design an algorithm that correctly counts and finds the\nreal zeros of polynomial systems that are located in the unbounded components\nof the complement of the underlying A-discriminant amoeba.\n", " The operational approach to the measurement of phase studied by Noh, Fougeres\nand Mandel is applied to the measurement of the state of polarization of fully\npolarized light. Operational counterparts of the quantum Stokes parameters are\nintroduced and their fluctuations are examined. It is shown that if the\npolarized field is weak, the measured fluctuations are influenced not only by\nthe quantum properties of the source field but also that of the measurement.\nThis character is reflected on the measured probability distributions of the\nparameters of polarization, which are also investigated independently for the\nfully polarized coherent states and the Fock states as the initial field\nstrength is varied. Finally, connection between the operational approach to the\nmeasurement of polarization and the su(2) interferometry is examined.\n", " We consider the problem of continuous quantum measurement of coherent\noscillations between two quantum states of an individual two-state system. It\nis shown that the interplay between the information acquisition and the\nbackaction dephasing of the oscillations by the detector imposes a fundamental\nlimit, equal to 4, on the signal-to-noise ratio of the measurement. The limit\nis universal, e.g., independent of the coupling strength between the detector\nand system, and results from the tendency of quantum measurement to localize\nthe system in one of the measured eigenstates.\n", " Frames have established themselves as a means to derive redundant, yet stable\ndecompositions of a signal for analysis or transmission, while also promoting\nsparse expansions. However, when the signal dimension is large, the computation\nof the frame measurements of a signal typically requires a large number of\nadditions and multiplications, and this makes a frame decomposition intractable\nin applications with limited computing budget. To address this problem, in this\npaper, we focus on frames in finite-dimensional Hilbert spaces and introduce\nsparsity for such frames as a new paradigm. In our terminology, a sparse frame\nis a frame whose elements have a sparse representation in an orthonormal basis,\nthereby enabling low-complexity frame decompositions. To introduce a precise\nmeaning of optimality, we take the sum of the numbers of vectors needed of this\northonormal basis when expanding each frame vector as sparsity measure. We then\nanalyze the recently introduced algorithm Spectral Tetris for construction of\nunit norm tight frames and prove that the tight frames generated by this\nalgorithm are in fact optimally sparse with respect to the standard unit vector\nbasis. Finally, we show that even the generalization of Spectral Tetris for the\nconstruction of unit norm frames associated with a given frame operator\nproduces optimally sparse frames.\n", " This paper is motivated by the k-nearest neighbors search: given an arbitrary\nmetric space, and its finite subsets (a reference set R and a query set Q),\ndesign a fast algorithm to find all k-nearest neighbors in R for every point q\nin Q. In 2006, Beygelzimer, Kakade, and Langford introduced cover trees to\njustify a near-linear time complexity for the neighbor search in the sizes of\nQ,R.\n Section 5.3 of Curtin's PhD (2015) pointed out that the proof of this result\nwas wrong. The key step in the original proof attempted to show that the number\nof iterations can be estimated by multiplying the length of the longest\nroot-to-leaf path in a cover tree by a constant factor. However, this estimate\ncan miss many potential nodes in several branches of a cover tree, that should\nbe considered during the neighbor search. The same argument was unfortunately\nrepeated in several subsequent papers using cover trees from 2006.\n This paper explicitly constructs challenging datasets that provide\ncounterexamples to the past proofs of time complexity for the cover tree\nconstruction, the k-nearest neighbor search presented at ICML 2006, and the\ndual-tree search algorithm published in NIPS 2009.\n The corrected near-linear time complexities with extra parameters are proved\nin another forthcoming paper by using a new compressed cover tree simplifying\nthe original tree structure.\n", " A 1965 problem due to Danzer asks whether there exists a set in Euclidean\nspace with finite density intersecting any convex body of volume one. A recent\napproach to this problem is concerned with the construction of dense forests\nand is obtained by a suitable weakening of the volume constraint. A dense\nforest is a discrete point set of finite density getting uniformly close to\nlong enough line segments. The distribution of points in a dense forest is then\nquantified in terms of a visibility function. Another way to weaken the\nassumptions in Danzer's problem is by relaxing the density constraint. In this\nrespect, a new concept is introduced in this paper, namely that of an optical\nforest. An optical forest in $\\mathbb{R}^{d}$ is a point set with optimal\nvisibility but not necessarily with finite density. In the literature, the best\nconstructions of Danzer sets and dense forests lack effectivity. The goal of\nthis paper is to provide constructions of dense and optical forests which yield\nthe best known results in any dimension $d \\ge 2$ both in terms of visibility\nand density bounds and effectiveness. Namely, there are three main results in\nthis work: (1) the construction of a dense forest with the best known\nvisibility bound which, furthermore, enjoys the property of being\ndeterministic; (2) the deterministic construction of an optical forest with a\ndensity failing to be finite only up to a logarithm and (3) the construction of\na planar Peres-type forest (that is, a dense forest obtained from a\nconstruction due to Peres) with the best known visibility bound. This is\nachieved by constructing a deterministic digital sequence satisfying strong\ndispersion properties.\n", " We present an approach to inform the reconstruction of a surface from a point\nscan through topological priors. The reconstruction is based on basis functions\nwhich are optimized to provide a good fit to the point scan while satisfying\npredefined topological constraints. We optimize the parameters of a model to\nobtain likelihood function over the reconstruction domain. The topological\nconstraints are captured by persistence diagrams which are incorporated in the\noptimization algorithm promote the correct topology. The result is a novel\ntopology-aware technique which can: 1.) weed out topological noise from point\nscans, and 2.) capture certain nuanced properties of the underlying shape which\ncould otherwise be lost while performing surface reconstruction. We showcase\nresults reconstructing shapes with multiple potential topologies, compare to\nother classical surface construction techniques, and show the completion of\nreal scan data.\n", " We show that the method in recent work of Roncal, Shrivastava, and Shuin can\nbe adapted to show that certain $L^p$-improving bounds in the interior of the\nboundedness region for the bilinear spherical or triangle averaging operator\nimply sparse bounds for the corresponding lacunary maximal operator, and that\n$L^p$-improving bounds in the interior of the boundedness region for the\ncorresponding single-scale maximal operators imply sparse bounds for the\ncorrepsonding full maximal operators. More generally we show that the framework\napplies for bilinear convolutions with compactly supported finite Borel\nmeasures that satisfy appropriate $L^p$-improving and continuity estimates.\nThis shows that the method used by Roncal, Shrivastava, and Shuin can be\nadapted to obtain sparse bounds for a general class of bilinear operators that\nare not of product type, for a certain range of $L^p$ exponents.\n", " We proposed a new criterion \\textit{noise-stability}, which revised the\nclassical rigidity theory, for evaluation of MDS algorithms which can\ntruthfully represent the fidelity of global structure reconstruction; then we\nproved the noise-stability of the cMDS algorithm in generic conditions, which\nprovides a rigorous theoretical guarantee for the precision and theoretical\nbounds for Euclidean embedding and its application in fields including wireless\nsensor network localization and satellite positioning.\n Furthermore, we looked into previous work about minimum-cost globally rigid\nspanning subgraph, and proposed an algorithm to construct a minimum-cost\nnoise-stable spanning graph in the Euclidean space, which enabled reliable\nlocalization on sparse graphs of noisy distance constraints with linear numbers\nof edges and sublinear costs in total edge lengths. Additionally, this\nalgorithm also suggests a scheme to reconstruct point clouds from pairwise\ndistances at a minimum of $O(n)$ time complexity, down from $O(n^3)$ for cMDS.\n", " In this paper, we consider the problem of improving the inference latency of\nlanguage model-based dense retrieval systems by introducing structural\ncompression and model size asymmetry between the context and query encoders.\nFirst, we investigate the impact of pre and post-training compression on the\nMSMARCO, Natural Questions, TriviaQA, SQUAD, and SCIFACT, finding that\nasymmetry in the dual encoders in dense retrieval can lead to improved\ninference efficiency. Knowing this, we introduce Kullback Leibler Alignment of\nEmbeddings (KALE), an efficient and accurate method for increasing the\ninference efficiency of dense retrieval methods by pruning and aligning the\nquery encoder after training. Specifically, KALE extends traditional Knowledge\nDistillation after bi-encoder training, allowing for effective query encoder\ncompression without full retraining or index generation. Using KALE and\nasymmetric training, we can generate models which exceed the performance of\nDistilBERT despite having 3x faster inference.\n", " It is widely known in the machine learning community that class noise can be\n(and often is) detrimental to inducing a model of the data. Many current\napproaches use a single, often biased, measurement to determine if an instance\nis noisy. A biased measure may work well on certain data sets, but it can also\nbe less effective on a broader set of data sets. In this paper, we present\nnoise identification using classifier diversity (NICD) -- a method for deriving\na less biased noise measurement and integrating it into the learning process.\nTo lessen the bias of the noise measure, NICD selects a diverse set of\nclassifiers (based on their predictions of novel instances) to determine which\ninstances are noisy. We examine NICD as a technique for filtering, instance\nweighting, and selecting the base classifiers of a voting ensemble. We compare\nNICD with several other noise handling techniques that do not consider\nclassifier diversity on a set of 54 data sets and 5 learning algorithms. NICD\nsignificantly increases the classification accuracy over the other considered\napproaches and is effective across a broad set of data sets and learning\nalgorithms.\n", " Repeated quantum measurements can generate effective new non-equilibrium\ndynamics in matter. Here we combine such a measurement driven system with\ndisorder. In particular, we investigate the diffusive behavior in the system\nand the effect of various types of disorder on the measurement induced chiral\ntransport protocol [1]. We begin by characterizing the diffusive behavior\nproduced by the measurements themselves in a clean system. We then examine the\nedge flow of particles per measurement cycle for three different types of\ndisorder: site dilution, lattice distortion, and disorder in onsite chemical\npotential. In the quantum Zeno limit, the effective descriptions for the\ndisordered measurement system with lattice distortions and random onsite\npotential can be modelled as a classical stochastic model, and the overall\neffect of increasing these disorders induces a crossover from perfect flow to\nzero transport. On the other hand if vacancies are present in the lattice the\nflow of particles per measurement cycle undergoes a percolation phase\ntransition from unity to zero with percolation threshold $p_c \\approx 0.26$,\nwith critical exponent $\\nu \\approx 1.35$. We also present numerical results\naway from Zeno limit and note that the overall effect of moving away from the\nZeno effect is to reduce particle flow per cycle when the measurement frequency\nin our protocol is reduced.\n", " A theorem of Bogolyubov states that for every dense set $A$ in $\\mathbb{Z}_N$\nwe may find a large Bohr set inside $A+A-A-A$. In this note, motivated by the\nwork on a quantitative inverse theorem for the Gowers $U^4$ norm, we prove a\nbilinear variant of this result in vector spaces over finite fields. Namely, if\nwe start with a dense set $A \\subset \\mathbb{F}^n_p \\times \\mathbb{F}^n_p$ and\nthen take rows (respectively columns) of $A$ and change each row (respectively\ncolumn) to the set difference of it with itself, repeating this procedure\nseveral times, we obtain a bilinear analogue of a Bohr set inside the resulting\nset, namely the zero set of a biaffine map from $\\mathbb{F}^n_p \\times\n\\mathbb{F}^n_p$ to a $\\mathbb{F}_p$-vector space of bounded dimension. An\nalmost identical result was proved independently by Bienvenu and L\\^e.\n", " Accessing fluid infiltration in nanogranular coatings is an outstanding\nchallenge, of relevance for applications ranging from nanomedicine to\ncatalysis. A sensing platform, allowing to quantify the amount of fluid\ninfiltrated in a nanogranular ultrathin coating, with thickness in the 10 to 40\nnm range, is here proposed and theoretically investigated by multiscale\nmodelling. The scheme relies on impulsive photoacoustic excitation of\nhypersonic mechanical breathing modes in engineered gas-phase synthesised\nnanogranular metallic ultathin films and time-resolved acousto-optical read-out\nof the breathing modes frequency shift upon liquid infiltration. A superior\nsensitivity, exceeding 26x103 cm^2/g, is predicted upon equivalent areal mass\nloading of a few ng/mm^2. The capability of the present scheme to discriminate\namong different infiltration patterns is discussed. The platform is an ideal\ntool to investigate nano fluidics in granular materials and naturally serves as\na distributed nanogetter coating, integrating fluid sensing capabilities. The\nproposed scheme is readily extendable to other nanoscale and mesoscale porous\nmaterials.\n", " We consider computing the Riemann zeta function $\\zeta(s)$ and Dirichlet\n$L$-functions $L(s,\\chi)$ to $p$-bit accuracy for large $p$. Using the\napproximate functional equation together with asymptotically fast computation\nof the incomplete gamma function, we observe that $p^{3/2+o(1)}$ bit complexity\ncan be achieved if $s$ is an algebraic number of fixed degree and with\nalgebraic height bounded by $O(p)$. This is an improvement over the\n$p^{2+o(1)}$ complexity of previously published algorithms and yields, among\nother things, $p^{3/2+o(1)}$ complexity algorithms for Stieltjes constants and\n$n^{3/2+o(1)}$ complexity algorithms for computing the $n$th Bernoulli number\nor the $n$th Euler number exactly.\n", " In this paper, we tested several sparse optimization algorithms based on the\npublic dataset of the DREAM5 Gene Regulatory Network Inference Challenge. And\nwe find that introducing 20% of the regulatory network as a priori known data\ncan provide a basis for parameter selection of inference algorithms, thus\nimproving prediction efficiency and accuracy. In addition to testing common\nsparse optimization methods, we also developed voting algorithms by bagging\nthem. Experiments on the DREAM5 dataset show that the sparse optimization-based\ninference of the moderation relation works well, achieving better results than\nthe official DREAM5 results on three datasets. However, the performance of\ntraditional independent algorithms varies greatly in the face of different\ndatasets, while our voting algorithm achieves the best results on three of the\nfour datasets.\n", " We describe and analyze an algorithm for computing the homology (Betti\nnumbers and torsion coefficients) of closed semialgebraic sets given by Boolean\nformulas without negations over lax polynomial inequalities. The algorithm\nworks in weak exponential time. This means that outside a subset of data having\nexponentially small measure, the cost of the algorithm is single exponential in\nthe size of the data.\n All previous algorithms solving this problem have doubly exponential\ncomplexity (and this is so for almost all input data). Our algorithm thus\nrepresents an exponential acceleration over state-of-the-art algorithms for all\ninput data outside a set that vanishes exponentially fast.\n", "4z(Assadi, Chen, and Khanna, 2019) define a 4-player hidden-pointer-chasing\n($\\mathsf{HPC}^4$), and using it, give strong multi-pass lower bounds for graph\nproblems in the streaming model of computation and a lower bound on the query\ncomplexity of sub-modular minimization. We present a two-player version\n($\\mathsf{HPC}^2$) of $\\mathsf{HPC}^4$ that has matching communication\ncomplexity to $\\mathsf{HPC}^4$. Our formulation allows us to lower bound its\ncommunication complexity with a simple direct-sum argument. Using this lower\nbound on the communication complexity of $\\mathsf{HPC}^2$, we retain the\nstreaming and query complexity lower bounds by (Assadi, Chen, and Khanna,\n2019).\n Further, by giving reductions from $\\mathsf{HPC}^2$, we prove new multi-pass\nspace lower bounds for graph problems in turnstile streams. In particular, we\nshow that any algorithm which computes the exact weight of the maximum weighted\nmatching in an $n$-vertex graph requires $\\tilde{O}(n^{2})$ space unless it\nmakes $\\omega(\\log n)$ passes over the turnstile stream, and that any algorithm\nwhich computes the minimum $s\\text{-}t$ distance in an $n$-vertex graph\nrequires $n^{2-o(1)}$ space unless it makes $n^{\\Omega(1)}$ passes over the\nturnstile stream. Our reductions can be modified to use $\\mathsf{HPC}^4$ as\nwell.\n", " The computation of determinants or their signs is the core procedure in many\nimportant geometric algorithms, such as convex hull, volume and point location.\nAs the dimension of the computation space grows, a higher percentage of the\ntotal computation time is consumed by these computations. In this paper we\nstudy the sequences of determinants that appear in geometric algorithms. The\ncomputation of a single determinant is accelerated by using the information\nfrom the previous computations in that sequence.\n We propose two dynamic determinant algorithms with quadratic arithmetic\ncomplexity when employed in convex hull and volume computations, and with\nlinear arithmetic complexity when used in point location problems. We implement\nthe proposed algorithms and perform an extensive experimental analysis. On one\nhand, our analysis serves as a performance study of state-of-the-art\ndeterminant algorithms and implementations. On the other hand, we demonstrate\nthe supremacy of our methods over state-of-the-art implementations of\ndeterminant and geometric algorithms. Our experimental results include a 20 and\n78 times speed-up in volume and point location computations in dimension 6 and\n11 respectively.\n", " Sparseness is a useful regularizer for learning in a wide range of\napplications, in particular in neural networks. This paper proposes a model\ntargeted at classification tasks, where sparse activity and sparse connectivity\nare used to enhance classification capabilities. The tool for achieving this is\na sparseness-enforcing projection operator which finds the closest vector with\na pre-defined sparseness for any given vector. In the theoretical part of this\npaper, a comprehensive theory for such a projection is developed. In\nconclusion, it is shown that the projection is differentiable almost everywhere\nand can thus be implemented as a smooth neuronal transfer function. The entire\nmodel can hence be tuned end-to-end using gradient-based methods. Experiments\non the MNIST database of handwritten digits show that classification\nperformance can be boosted by sparse activity or sparse connectivity. With a\ncombination of both, performance can be significantly better compared to\nclassical non-sparse approaches.\n", " This work addresses the block-diagonal semidefinite program (SDP) relaxations\nfor the clique number of the Paley graphs. The size of the maximal clique\n(clique number) of a graph is a classic NP-complete problem; a Paley graph is a\ndeterministic graph where two vertices are connected if their difference is a\nquadratic residue (square) in a finite field with the number of elements given\nby certain primes and prime powers. Improving the upper bound for the Paley\ngraph clique number for prime powers that are non-squares is an open problem in\ncombinatorics. Moreover, since quadratic residues exhibit pseudorandom\nproperties, Paley graphs are related to the construction of deterministic\nrestricted isometries, an open problem in compressed sensing. Recent work\nprovides numerical evidence that the current upper bounds can be improved by\nthe sum-of-squares (SOS) relaxations. In particular, the bounds given by the\nSOS relaxations of degree 4 (SOS-4) have been empirically observed to be\ngrowing at an order smaller than square root of the prime. However,\ncomputations of SOS-4 appear to be intractable with respect to large graphs.\nGvozdenovic et al. introduced a more computationally efficient block-diagonal\nhierarchy of SDPs and computed the values of these SDPs of degrees 2 (L2) for\nthe Paley graph clique numbers associated with primes p less or equal to 809,\nwhich bound from above the corresponding SOS-4 relaxations. We compute the\nvalues of the L2 relaxations for p's between 821 and 997. Our results provide\nsome numerical evidence that these relaxations, and therefore also the SOS-4\nrelaxations, may be scaling at an order smaller than the square root of p.\nHowever, due to the size of the SDPs, we have not been able to compute L2\nrelaxations for p's greater than 997. Therefore, our scaling estimate is not\nconclusive and presents an interesting open problem for further study.\n", " Quantitatively predictive models of biomolecular circuits are important tools\nfor the design of synthetic biology and molecular communication circuits. The\ninformation content of typical time-lapse single-cell data for the inference of\nkinetic parameters is not only limited by measurement uncertainty and intrinsic\nstochasticity, but also by the employed perturbations. Novel microfluidic\ndevices enable the synthesis of temporal chemical concentration profiles. The\ninformativeness of a perturbation can be quantified based on mutual\ninformation. We propose an approximate method to perform optimal experimental\ndesign of such perturbation profiles. To estimate the mutual information we\nperform a multivariate log-normal approximation of the joint distribution over\nparameters and observations and scan the design space using Metropolis-Hastings\nsampling. The method is demonstrated by finding optimal perturbation sequences\nfor synthetic case studies on a gene expression model with varying reporter\ncharacteristics.\n", " Any organism is embedded in an environment that changes over time. The\ntimescale for and statistics of environmental change, the precision with which\nthe organism can detect its environment, and the costs and benefits of\nparticular protein expression levels all will affect the suitability of\ndifferent strategies-such as constitutive expression or graded response-for\nregulating protein levels in response to environmental inputs. We propose a\ngeneral framework-here specifically applied to the enzymatic regulation of\nmetabolism in response to changing concentrations of a basic nutrient-to\npredict the optimal regulatory strategy given the statistics of fluctuations in\nthe environment and measurement apparatus, respectively, and the costs\nassociated with enzyme production. We use this framework to address three\nfundamental questions: (i) when a cell should prefer thresholding to a graded\nresponse; (ii) when there is a fitness advantage to implementing a Bayesian\ndecision rule; and (iii) when retaining memory of the past provides a selective\nadvantage. We specifically find that: (i) relative convexity of enzyme\nexpression cost and benefit influences the fitness of thresholding or graded\nresponses; (ii) intermediate levels of measurement uncertainty call for a\nsophisticated Bayesian decision rule; and (iii) in dynamic contexts,\nintermediate levels of uncertainty call for retaining memory of the past.\nStatistical properties of the environment, such as variability and correlation\ntimes, set optimal biochemical parameters, such as thresholds and decay rates\nin signaling pathways. Our framework provides a theoretical basis for\ninterpreting molecular signal processing algorithms and a classification scheme\nthat organizes known regulatory strategies and may help conceptualize\nheretofore unknown ones.\n", " We study and compare the gap and the Riesz topologies of the space of all\nunbounded regular operators on Hilbert C*-modules. We show that the space of\nall bounded adjointable operators on Hilbert C*-modules is an open dense subset\nof the space of all unbounded regular operators with respect to the gap\ntopology. The restriction of the gap topology on the space of all bounded\nadjointable operators is equivalent with the topology which is generated by the\nusual operator norm. The space of regular selfadjoint Fredholm operators on\nHilbert C*-modules over the C*-algebra of compact operators is path-connected\nwith respect to the gap topology, however, the result may not be true for some\nHilbert C*-modules.\n", " We propose a new search strategy based on the event shape variables for new\nphysics models where the separations among the masses of the particles in the\nspectrum are small. Collider signature of these models, characterised by low\n$p_T$ leptons/jets and low missing $p_T$, are known to be difficult to look\nfor. The conventional search strategies involving hard cuts may not work in\nsuch situations. As a case study, we have investigated the hitherto neglected\njets + missing $E_T$ signature -known to be a challenging one - arising from\nthe pair productions and decay of $n =1$ KK-excitations of gluons and quarks in\nthe minimal Universal Extra Dimension (mUED) model. Judicious use of the event\nshape variables, enables us to reduce the Standard Model backgrounds to a\nnegligible level . We have shown that in mUED, $R^{-1}$ upto $850 ~\\rm GeV$,\ncan be explored or ruled out with 12 fb$^{-1}$ of integrated luminosity at the\n7 TeV run of the LHC. We also discuss the prospects of employing these\nvariables for searching other beyond standard model physics with compressed or\npartially compressed spectra.\n", " Seminal works on light spanners over the years provide spanners with optimal\nlightness in various graph classes, such as in general graphs, Euclidean\nspanners, and minor-free graphs. Three shortcomings of previous works on light\nspanners are: (1) The techniques are ad hoc per graph class, and thus can't be\napplied broadly. (2) The runtimes of these constructions are almost always\nsub-optimal, and usually far from optimal. (3) These constructions are optimal\nin the standard and crude sense, but not in a refined sense that takes into\naccount a wider range of involved parameters.\n This work aims at addressing these shortcomings by presenting a unified\nframework of light spanners in a variety of graph classes. Informally, the\nframework boils down to a transformation from sparse spanners to light\nspanners; since the state-of-the-art for sparse spanners is much more advanced\nthan that for light spanners, such a transformation is powerful. Our framework\nis developed in two papers. The current paper is the first of the two -- it\nlays the basis of the unified framework and then applies it to design fast\nconstructions with optimal lightness for several graph classes.\n Among various applications and implications of our framework, we highlight\nhere the following:\n _ In low-dimensional Euclidean spaces, we present an $O(n\\log n)$-time\nconstruction of $(1+\\epsilon)$-spanners with lightness and degree both bounded\nby constants in the algebraic computation tree (ACT). Our construction resolves\na major problem in the area of geometric spanners, which was open for three\ndecades.\n _ In general graphs, for any $k \\geq 2$, we construct a\n$(2k-1)(1+\\epsilon)$-spanner with lightness $O(n^{1/k})$ in $O(m \\alpha(m,n))$\ntime. This result for light spanners in general weighted graphs is surprising,\nas it outperforms the analog one for sparse spanners.\n", " The frequent misunderstanding of information entropy is pointed out. It is\nshown that, contrary to fortuitous situations and common beliefs, there is no\ngeneral interrelation between the information entropy and the thermodynamical\nentropy. We point out that the change of information entropy during measurement\nis determined by the resolution of the measurement instrument and these changes\ndo not have a general and clear-cut interrelation with the change of thermal\nentropy. Moreover, these changes can be separated in space and time. We show\nclassical physical considerations and examples to support this conclusion.\nFinally, we show that information entropy can violate the Third Law of\nThermodynamics which is another indication of major differences from thermal\nentropy.\n", " We analyse the unreduced, nonperturbative dynamics of an arbitrary many-body\ninteraction process with the help of the generalised effective potential method\nand reveal the well-specified universal origin of change (emergence), time and\nevolution in an a priori conservative, time-independent system. It appears\ntogether with the universal dynamic complexity definition, where this unified\ncomplexity conservation and transformation constitutes the essence of\nevolution. We then consider the detailed structure of this universal\nevolutionary process showing its step-wise, \"punctuated\" character, now\nprovided with the exact mathematical description. Comparing the expected\nfeatures of a revolutionary complexity transition near a step-like complexity\nupgrade with the currently observed behaviour of world's social and economic\nsystems, we prove the necessity of complexity revolution towards the superior\ncivilisation level of well-defined nature, the only alternative being an\nequally dramatic and irreversible degradation, irrespective of efforts applied\nto stop the crisis at the current totally saturated complexity level.\n", " Due to the dynamic nature of biological systems, biological networks\nunderlying temporal process such as the development of {\\it Drosophila\nmelanogaster} can exhibit significant topological changes to facilitate dynamic\nregulatory functions. Thus it is essential to develop methodologies that\ncapture the temporal evolution of networks, which make it possible to study the\ndriving forces underlying dynamic rewiring of gene regulation circuity, and to\npredict future network structures. Using a new machine learning method called\nTesla, which builds on a novel temporal logistic regression technique, we\nreport the first successful genome-wide reverse-engineering of the latent\nsequence of temporally rewiring gene networks over more than 4000 genes during\nthe life cycle of \\textit{Drosophila melanogaster}, given longitudinal gene\nexpression measurements and even when a single snapshot of such measurement\nresulted from each (time-specific) network is available. Our methods offer the\nfirst glimpse of time-specific snapshots and temporal evolution patterns of\ngene networks in a living organism during its full developmental course. The\nrecovered networks with this unprecedented resolution chart the onset and\nduration of many gene interactions which are missed by typical static network\nanalysis, and are suggestive of a wide array of other temporal behaviors of the\ngene network over time not noticed before.\n", " A class of graphs $\\mathscr{C}$ is monadically stable if for any unary\nexpansion $\\widehat{\\mathscr{C}}$ of $\\mathscr{C}$, one cannot interpret, in\nfirst-order logic, arbitrarily long linear orders in graphs from\n$\\widehat{\\mathscr{C}}$. It is known that nowhere dense graph classes are\nmonadically stable; these encompass most of the studied concepts of sparsity in\ngraphs, including graph classes that exclude a fixed topological minor. On the\nother hand, monadic stability is a property expressed in purely model-theoretic\nterms and hence it is also suited for capturing structure in dense graphs.\n For several years, it has been suspected that one can create a structure\ntheory for monadically stable graph classes that mirrors the theory of nowhere\ndense graph classes in the dense setting. In this work we provide a step in\nthis direction by giving a characterization of monadic stability through the\nFlipper game: a game on a graph played by Flipper, who in each round can\ncomplement the edge relation between any pair of vertex subsets, and Connector,\nwho in each round localizes the game to a ball of bounded radius. This is an\nanalog of the Splitter game, which characterizes nowhere dense classes of\ngraphs (Grohe, Kreutzer, and Siebertz, J.ACM'17).\n We give two different proofs of our main result. The first proof uses tools\nfrom model theory, and it exposes an additional property of monadically stable\ngraph classes that is close in spirit to definability of types. Also, as a\nbyproduct, we give an alternative proof of the recent result of Braunfeld and\nLaskowski (arXiv 2209.05120) that monadic stability for graph classes coincides\nwith existential monadic stability. The second proof relies on the recently\nintroduced notion of flip-wideness (Dreier, M\\\"ahlmann, Siebertz, and\nToru\\'nczyk, arXiv 2206.13765) and provides an efficient algorithm to compute\nFlipper's moves in a winning strategy.\n", " We introduce a definition of algorithmic symmetry able to capture essential\naspects of geometric symmetry. We review, study and apply a method for\napproximating the algorithmic complexity (also known as Kolmogorov-Chaitin\ncomplexity) of graphs and networks based on the concept of Algorithmic\nProbability (AP). AP is a concept (and method) capable of recursively\nenumeration all properties of computable (causal) nature beyond statistical\nregularities. We explore the connections of algorithmic complexity---both\ntheoretical and numerical---with geometric properties mainly symmetry and\ntopology from an (algorithmic) information-theoretic perspective. We show that\napproximations to algorithmic complexity by lossless compression and an\nAlgorithmic Probability-based method can characterize properties of\npolyominoes, polytopes, regular and quasi-regular polyhedra as well as\npolyhedral networks, thereby demonstrating its profiling capabilities.\n", " Convolutional Neural Networks (CNN) are the state-of-the-art in the field of\nvisual computing. However, a major problem with CNNs is the large number of\nfloating point operations (FLOPs) required to perform convolutions for large\ninputs. When considering the application of CNNs to video data, convolutional\nfilters become even more complex due to the extra temporal dimension. This\nleads to problems when respective applications are to be deployed on mobile\ndevices, such as smart phones, tablets, micro-controllers or similar,\nindicating less computational power.\n Kim et al. (2016) proposed using a Tucker-decomposition to compress the\nconvolutional kernel of a pre-trained network for images in order to reduce the\ncomplexity of the network, i.e. the number of FLOPs. In this paper, we\ngeneralize the aforementioned method for application to videos (and other 3D\nsignals) and evaluate the proposed method on a modified version of the THETIS\ndata set, which contains videos of individuals performing tennis shots. We show\nthat the compressed network reaches comparable accuracy, while indicating a\nmemory compression by a factor of 51. However, the actual computational\nspeed-up (factor 1.4) does not meet our theoretically derived expectation\n(factor 6).\n", " In this paper we investigate the $L^p$ boundedness of the lacunary maximal\nfunction $ M_{\\Ha}^{lac} $ associated to the spherical means $ A_r f$ taken\nover Koranyi spheres on the Heisenberg group. Closely following an approach\nused by M. Lacey in the Euclidean case, we obtain sparse bounds for these\nmaximal functions leading to new unweighted and weighted estimates. The key\ningredients in the proof are the $L^p$ improving property of the operator\n$A_rf$ and a continuity property of the difference $A_rf-\\tau_y A_rf$, where\n$\\tau_yf(x)=f(xy^{-1})$ is the right translation operator.\n", " Network Pruning is a promising way to address the huge computing resource\ndemands of the deployment and inference of Large Language Models (LLMs).\nRetraining-free is important for LLMs' pruning methods. However, almost all of\nthe existing retraining-free pruning approaches for LLMs focus on unstructured\npruning, which requires specific hardware support for acceleration. In this\npaper, we propose a novel retraining-free structured pruning framework for\nLLMs, named FLAP (FLuctuation-based Adaptive Structured Pruning). It is\nhardware-friendly by effectively reducing storage and enhancing inference\nspeed. For effective structured pruning of LLMs, we highlight three critical\nelements that demand the utmost attention: formulating structured importance\nmetrics, adaptively searching the global compressed model, and implementing\ncompensation mechanisms to mitigate performance loss. First, FLAP determines\nwhether the output feature map is easily recoverable when a column of weight is\nremoved, based on the fluctuation pruning metric. Then it standardizes the\nimportance scores to adaptively determine the global compressed model\nstructure. At last, FLAP adds additional bias terms to recover the output\nfeature maps using the baseline values. We thoroughly evaluate our approach on\na variety of language benchmarks. Without any retraining, our method\nsignificantly outperforms the state-of-the-art methods, including LLM-Pruner\nand the extension of Wanda in structured pruning. The code is released at\nhttps://github.com/CASIA-IVA-Lab/FLAP.\n" ]

[ " Geometric disentanglement, the separation of latent codes for intrinsic (i.e.\nidentity) and extrinsic(i.e. pose) geometry, is a prominent task for generative\nmodels of non-Euclidean data such as 3D deformable models. It provides greater\ninterpretability of the latent space, and leads to more control in generation.\nThis work introduces a mesh feature, the conformal factor and normal feature\n(CFAN),for use in mesh convolutional autoencoders. We further propose CFAN-VAE,\na novel architecture that disentangles identity and pose using the CFAN\nfeature. Requiring no label information on the identity or pose during\ntraining, CFAN-VAE achieves geometric disentanglement in an unsupervisedway.\nOur comprehensive experiments, including reconstruction, interpolation,\ngeneration, and identity/pose transfer, demonstrate CFAN-VAE achieves\nstate-of-the-art performance on unsupervised geometric disentanglement. We also\nsuccessfully detect a level of geometric disentanglement in mesh convolutional\nautoencoders that encode xyz-coordinates directly by registering its latent\nspace to that of CFAN-VAE.\n", " We construct bijections giving three \"codes\" for trees. These codes follow\nnaturally from the Matrix Tree Theorem of Tutte and have many advantages over\nthe one produced by Prufer in 1918. One algorithm gives explicitly a bijection\nthat is implicit in Orlin's manipulatorial proof of Cayley's formula (the\nformula was actually found first by Borchardt). Another is based on a proof of\nKnuth. The third is an implementation of Joyal's pseudo-bijective proof of the\nformula, and is equivalent to one previously found by Egecioglu and Remmel. In\neach case, we have at least two algorithms, one of which involves hands-on\nmanipulations of the tree while the other involves a combinatorial and linear\nalgebraic manipulation of a matrix.\n", " We propose a partially learned approach for the solution of ill posed inverse\nproblems with not necessarily linear forward operators. The method builds on\nideas from classical regularization theory and recent advances in deep learning\nto perform learning while making use of prior information about the inverse\nproblem encoded in the forward operator, noise model and a regularizing\nfunctional. The method results in a gradient-like iterative scheme, where the\n\"gradient\" component is learned using a convolutional network that includes the\ngradients of the data discrepancy and regularizer as input in each iteration.\nWe present results of such a partially learned gradient scheme on a non-linear\ntomographic inversion problem with simulated data from both the Sheep-Logan\nphantom as well as a head CT. The outcome is compared against FBP and TV\nreconstruction and the proposed method provides a 5.4 dB PSNR improvement over\nthe TV reconstruction while being significantly faster, giving reconstructions\nof 512 x 512 volumes in about 0.4 seconds using a single GPU.\n", " In this paper we characterize the algebraic structure of two-dimensional\n$(\\alpha,\\beta )$-constacyclic codes of arbitrary length $s.\\ell$ and of their\nduals. For $\\alpha,\\beta \\in \\{1,-1\\}$, we give necessary and sufficient\nconditions for a two-dimensional $(\\alpha,\\beta )$-constacyclic code to be\nself-dual. We also show that a two-dimensional $(\\alpha,1 )$-constacyclic code\n$\\mathcal{C}$ of length $n=s.\\ell$ can not be self-dual if $\\gcd(s,q)= 1$.\nFinally, we give some examples of self-dual, isodual, MDS and quasi-twisted\ncodes corresponding to two-dimensional $(\\alpha,\\beta )$-constacyclic codes.\n", " In this article we present a class of codes with few weights arising from\nspecial type of linear sets. We explicitly show the weights of such codes,\ntheir weight enumerator and possible choices for their generator matrices. In\nparticular, our construction yields also to linear codes with three weights\nand, in some cases, to almost MDS codes. The interest for these codes relies on\ntheir applications to authentication codes and secret schemes, and their\nconnections with further objects such as association schemes and graphs.\n", " Many-fermion Hilbert space has the algebraic structure of a free module\ngenerated by a finite number of antisymmetric functions called shapes.\nPhysically, each shape is a many-body vacuum, whose excitations are described\nby symmetric functions (bosons). The infinity of bosonic excitations accounts\nfor the infinity of Hilbert space, while all shapes can be generated\nalgorithmically in closed form. The shapes are geometric objects in\nwave-function space, such that any given many-body vacuum is their\nintersection. Correlation effects in laboratory space are geometric constraints\nin wave-function space. Algebraic geometry is the natural mathematical\nframework for the particle picture of quantum mechanics. Simple examples of\nthis scheme are given, and the current state of the art in generating shapes is\ndescribed from the viewpoint of treating very large function spaces.\n", " We consider a mobile edge computing scenario where users want to perform a\nlinear inference operation $\\boldsymbol{W} \\boldsymbol{x}$ on local data\n$\\boldsymbol{x}$ for some network-side matrix $\\boldsymbol{W}$. The inference\nis performed in a distributed fashion over multiple servers at the network\nedge. For this scenario, we propose a coding scheme that combines a rateless\ncode to provide resiliency against straggling servers--hence reducing the\ncomputation latency--and an irregular-repetition code to provide spatial\ndiversity--hence reducing the communication latency. We further derive a lower\nbound on the total latency--comprising computation latency, communication\nlatency, and decoding latency. The proposed scheme performs remarkably close to\nthe bound and yields significantly lower latency than the scheme based on\nmaximum distance separable codes recently proposed by Zhang and Simeone.\n", " In diffusion MRI (dMRI), a good sampling scheme is important for efficient\nacquisition and robust reconstruction. Diffusion weighted signal is normally\nacquired on single or multiple shells in q-space. Signal samples are typically\ndistributed uniformly on different shells to make them invariant to the\norientation of structures within tissue, or the laboratory coordinate frame.\nThe Electrostatic Energy Minimization (EEM) method, originally proposed for\nsingle shell sampling scheme in dMRI, was recently generalized to multi-shell\nschemes, called Generalized EEM (GEEM). GEEM has been successfully used in the\nHuman Connectome Project (HCP). However, EEM does not directly address the goal\nof optimal sampling, i.e., achieving large angular separation between sampling\npoints. In this paper, we propose a more natural formulation, called Spherical\nCode (SC), to directly maximize the minimal angle between different samples in\nsingle or multiple shells. We consider not only continuous problems to design\nsingle or multiple shell sampling schemes, but also discrete problems to\nuniformly extract sub-sampled schemes from an existing single or multiple shell\nscheme, and to order samples in an existing scheme. We propose five algorithms\nto solve the above problems, including an incremental SC (ISC), a sophisticated\ngreedy algorithm called Iterative Maximum Overlap Construction (IMOC), an 1-Opt\ngreedy method, a Mixed Integer Linear Programming (MILP) method, and a\nConstrained Non-Linear Optimization (CNLO) method. To our knowledge, this is\nthe first work to use the SC formulation for single or multiple shell sampling\nschemes in dMRI. Experimental results indicate that SC methods obtain larger\nangular separation and better rotational invariance than the state-of-the-art\nEEM and GEEM. The related codes and a tutorial have been released in DMRITool.\n", " The aim of this article is to establish two-sided Gaussian bounds for the\nheat kernels on the unit ball and simplex in $\\mathbb{R}^n$, and in particular\non the interval, generated by classical differential operators whose\neigenfunctions are algebraic polynomials. To this end we develop a general\nmethod that employs the natural relation of such operators with weighted\nLaplace operators on suitable subsets of Riemannian manifolds and the existing\ngeneral results on heat kernels. Our general scheme allows to consider heat\nkernels in the weighted cases on the interval, ball, and simplex with\nparameters in the full range.\n", " In this paper we analyse the construction of identification codes.\nIdentification codes are based on the question \"Is the message I have just\nreceived the one I am interested in?\", as opposed to Shannon's transmission,\nwhere the receiver is interested in not only one, but any message. The\nadvantage of identification is that it allows rates growing double\nexponentially in the blocklength at the cost of not being able to decode every\nmessage, which might be beneficial in certain applications. We focus on a\nspecial identification code construction based on two concatenated Reed-Solomon\ncodes and have a closer look at its implementation, analyzing the trade-offs of\nidentification with respect to transmission and the trade-offs introduced by\nthe computational cost of identification codes.\n", " Let $\\mathcal{C}$ be a set of $m$ by $n$ matrices over $\\mathbb{F}_q$ such\nthat the rank of $A-B$ is at least $d$ for all distinct $A,B\\in \\mathcal{C}$.\nSuppose that $m\\leqslant n$. If $\\#\\mathcal{C}= q^{n(m-d+1)}$, then\n$\\mathcal{C}$ is a maximum rank distance (MRD for short) code. Until 2016,\nthere were only two known constructions of MRD codes for arbitrary $1<d<m-1$.\nOne was found by Delsarte (1978) and Gabidulin (1985) independently, and it was\nlater generalized by Kshevetskiy and Gabidulin (2005). We often call them\n(generalized) Gabidulin codes. Another family was recently obtained by Sheekey\n(2016), and its elements are called twisted Gabidulin codes. In the same paper,\nSheekey also proposed a generalization of the twisted Gabidulin codes. However\nthe equivalence problem for it is not considered, whence it is not clear\nwhether there exist new MRD codes in this generalization. We call the members\nof this putative larger family generalized twisted Gabidulin codes. In this\npaper, we first compute the Delsarte duals and adjoint codes of them, then we\ncompletely determine the equivalence between different generalized twisted\nGabidulin codes. In particular, it can be proven that, up to equivalence,\ngeneralized Gabidulin codes and twisted Gabidulin codes are both proper subsets\nof this family.\n", " The exploration of linear subspaces, particularly scattered subspaces, has\ngarnered considerable attention across diverse mathematical disciplines in\nrecent years, notably within finite geometries and coding theory. Scattered\nsubspaces play a pivotal role in analyzing various geometric structures such as\nblocking sets, two-intersection sets, complete arcs, caps in affine and\nprojective spaces over finite fields and rank metric codes. This paper\nintroduces a new infinite family of $h$-subspaces, along with their associated\nMRD codes. Additionally, it addresses the task of determining the generalized\nweights of these codes. Notably, we demonstrate that these MRD codes exhibit\nsome larger generalized weights compared to those previously identified.\n", " Convolutional Neural Networks have been used in a variety of image related\napplications after their rise in popularity due to ImageNet competition.\nConvolutional Neural Networks have shown remarkable results in applications\nincluding face recognition, moving target detection and tracking,\nclassification of food based on the calorie content and many more. Designing of\nConvolutional Neural Networks requires experts having a cross domain knowledge\nand it is laborious, which requires a lot of time for testing different values\nfor different hyperparameter along with the consideration of different\nconfigurations of existing architectures. Neural Architecture Search is an\nautomated way of generating Neural Network architectures which saves\nresearchers from all the brute-force testing trouble, but with the drawback of\nconsuming a lot of computational resources for a prolonged period. In this\npaper, we propose an automated Neural Architecture Search framework DQNAS,\nguided by the principles of Reinforcement Learning along with One-shot Training\nwhich aims to generate neural network architectures that show superior\nperformance and have minimum scalability problem.\n", " The aim of this work is to give degree formulas for the generalized Hamming\nweights of evaluation codes and to show lower bounds for these weights. In\nparticular, we give degree formulas for the generalized Hamming weights of\nReed--Muller-type codes, and we determine the minimum distance of toric codes\nover hypersimplices, and the 1st and 2nd generalized Hamming weights of\nsquarefree evaluation codes.\n", " In this paper we give a partial shift version of user-irrepressible sequence\nsets and conflict-avoiding codes. By means of disjoint difference sets, we\nobtain an infinite number of such user-irrepressible sequence sets whose\nlengths are shorter than known results in general. Subsequently, the newly\ndefined partially conflict-avoiding codes are discussed.\n", " In this paper, we define Abelian and consta-Abelian polyadic codes over rings\ndefined as affine algebras over chain rings. For that aim, we use the classical\nconstruction via splittings and multipliers of the underlying Abelian group. We\nalso derive some results on the structure of the associated polyadic codes and\nthe number of codes under these conditions.\n", " An alternative proof is given of the existence of greatest lower bounds in\nthe imbalance order of binary maximal instantaneous codes of a given size.\nThese codes are viewed as maximal antichains of a given size in the infinite\nbinary tree of 0-1 words. The proof proposed makes use of a single balancing\noperation instead of expansion and contraction as in the original proof of the\nexistence of glb.\n", " In the present paper we introduce and study finite point subsets of a special\nkind, called optimum distributions, in the n-dimensional unit cube. Such\ndistributions are closely related with known (delta,s,n)-nets of low\ndiscrepancy. It turns out that optimum distributions have a rich combinatorial\nstructure. Namely, we show that optimum distributions can be characterized\ncompletely as maximum distance separable codes with respect to a non-Hamming\nmetric. Weight spectra of such codes can be evaluated precisely. We also\nconsider linear codes and distributions and study their general properties\nincluding the duality with respect to a suitable inner product. The\ncorresponding generalized MacWilliams identities for weight enumerators are\nbriefly discussed. Broad classes of linear maximum distance separable codes and\nlinear optimum distributions are explicitely constructed in the paper by the\nHermite interpolations over finite fields.\n", " Neural codes serve as a language for neurons in the brain. Convex codes,\nwhich arise from the pattern of intersections of convex sets in Euclidean\nspace, are of particular relevance to neuroscience. Not every code is convex,\nhowever, and the combinatorial properties of a code that determine its\nconvexity are still poorly understood. Here we find that a code that can be\nrealized by a collection of open convex sets may or may not be realizable by\nclosed convex sets, and vice versa, establishing that open convex and closed\nconvex codes are distinct classes. We also prove that max intersection-complete\ncodes (i.e. codes that contain all intersections of maximal codewords) are both\nopen convex and closed convex, and provide an upper bound for their minimal\nembedding dimension. Finally, we show that the addition of non-maximal\ncodewords to an open convex code preserves convexity.\n", " Huffman Codes are optimal Instantaneous Fixed-to-Variable (FV) codes in which\nevery source symbol can only be encoded by one codeword. Relaxing these\nconstraints permits constructing better FV codes. More specifically, recent\nwork has shown that AIFV-$m$ codes can beat Huffman coding. AIFV-$m$ codes\nconstruct am $m$-tuple of different coding trees between which the code\nalternates and are only almost instantaneous (AI). This means that decoding a\nword might require a delay of a finite number of bits.\n Current algorithms for constructing optimal AIFV-$m$ codes are iterative\nprocesses that construct progressively \"better sets\" of code trees. The\nprocesses have been proven to finitely converge to the optimal code but with no\nknown bounds on the convergence rate.\n This paper derives a geometric interpretation of the space of binary AIFV-$2$\ncodes, permitting the development of the first polynomially time-bounded\niterative procedures for constructing optimal AIFV codes.\n We first show that a simple binary search procedure can replace the current\niterative process to construct optimal binary AIFV-$2$ codes. We then describe\nhow to frame the problem as a linear programming with an exponential number of\nconstraints but a polynomial-time separability oracle. This permits using the\nGr\\\"otschel, Lov\\'asz and Schrijver ellipsoid method to solve the problem in a\npolynomial number of steps. While more complicated, this second method has the\npotential to lead to a polynomial time algorithm to construct optimal AIFV-$m$\ncodes for general $m$.\n", " We discuss, in terms of rate-distortion theory, the fitness of molecular\ncodes as the problem of designing an optimal information channel. The fitness\nis governed by an interplay between the cost and quality of the channel, which\ninduces smoothness in the code. By incorporating this code fitness into\npopulation dynamics models, we suggest that the emergence and evolution of\nmolecular codes may be explained by simple channel design considerations.\n", " This paper concerns an analytical stratification question of real algebraic\nand semi-algebraic sets. For Whitney's stratification in 1957, it partitions a\nreal algebraic set into partial algebraic manifolds\\cite{W}. In 1975 Hironaka\nreproved that a real algebraic set is triangulable and also generalized it to\nsemi-algebraic sets, following the idea of Lojasiewicz's triangulation of\nsemi-analytic sets in 1964. Following their examples and wondering how geometry\nlooks like locally. this paper tries to come up with a stratification, in\nparticular a cell decomposition, such that it satisfies the following\nanalytical property. Given any shortest curve between two points in a real\nalgebraic or semi-algebraic set, it interacts each cell (or simplex) at most\nfinitely many times.\n", " Using the concept of projective systems for linear codes and elementary\nlinear algebra, we show that projective $[n,k]_q$ codes form a connected\nsubgraph in the Grassmann graph consisting of $k$-dimensional subspaces of an\n$n$-dimensional vector space over the $q$-element field.\n", " Convolutional Neural Networks (CNN) are the state-of-the-art in the field of\nvisual computing. However, a major problem with CNNs is the large number of\nfloating point operations (FLOPs) required to perform convolutions for large\ninputs. When considering the application of CNNs to video data, convolutional\nfilters become even more complex due to the extra temporal dimension. This\nleads to problems when respective applications are to be deployed on mobile\ndevices, such as smart phones, tablets, micro-controllers or similar,\nindicating less computational power.\n Kim et al. (2016) proposed using a Tucker-decomposition to compress the\nconvolutional kernel of a pre-trained network for images in order to reduce the\ncomplexity of the network, i.e. the number of FLOPs. In this paper, we\ngeneralize the aforementioned method for application to videos (and other 3D\nsignals) and evaluate the proposed method on a modified version of the THETIS\ndata set, which contains videos of individuals performing tennis shots. We show\nthat the compressed network reaches comparable accuracy, while indicating a\nmemory compression by a factor of 51. However, the actual computational\nspeed-up (factor 1.4) does not meet our theoretically derived expectation\n(factor 6).\n", " In this paper, motivated by the work of Edelman and Strang, we show that for\nfixed integers $d\\geq 2$ and $n\\geq d+1$ the configuration space of all facet\nvolume vectors of all $d$-polytopes in $\\mathbb R^{d}$ with $n$ facets is a\nfull dimensional cone in $\\mathbb R^{n}$. In particular, for tetrahedra ($d=3$\nand $n=4$) this is a cone over a regular octahedron. Our proof is based on a\nnovel configuration space / test map scheme which uses topological methods for\nfinding solutions of a problem, and tools of differential geometry to identify\nsolutions with the desired properties. Furthermore, our results open a\npossibility for the study of realization spaces of all $d$-polytopes in\n$\\mathbb R^{d}$ with $n$ facets by the methods of algebraic topology.\n", " A set $C$ of unit vectors in $\\mathbb{R}^d$ is called an $L$-spherical code\nif $x \\cdot y \\in L$ for any distinct $x,y$ in $C$. Spherical codes have been\nextensively studied since their introduction in the 1970's by Delsarte,\nGoethals and Seidel. In this note we prove a conjecture of Bukh on the maximum\nsize of spherical codes. In particular, we show that for any set of $k$ fixed\nangles, one can choose at most $O(d^k)$ lines in $\\mathbb{R}^d$ such that any\npair of them forms one of these angles.\n", " A computational strategy and a collection of codes are presented for studying\nmultiparticle final states in hard hadronic collisions.\n", " Existing approaches to the crime prediction problem are unsuccessful in\nexpressing the details since they assign the probability values to large\nregions. This paper introduces a new architecture with the graph convolutional\nnetworks (GCN) and multivariate Gaussian distributions to perform\nhigh-resolution forecasting that applies to any spatiotemporal data. We tackle\nthe sparsity problem in high resolution by leveraging the flexible structure of\nGCNs and providing a subdivision algorithm. We build our model with Graph\nConvolutional Gated Recurrent Units (Graph-ConvGRU) to learn spatial, temporal,\nand categorical relations. In each node of the graph, we learn a multivariate\nprobability distribution from the extracted features of GCNs. We perform\nexperiments on real-life and synthetic datasets, and our model obtains the best\nvalidation and the best test score among the baseline models with significant\nimprovements. We show that our model is not only generative but also precise.\n", " A new generalization of the Gray map is introduced. The new generalization\n$\\Phi: Z_{2^k}^n \\to Z_{2}^{2^{k-1}n}$ is connected with the known generalized\nGray map $\\phi$ in the following way: if we take two dual linear\n$Z_{2^k}$-codes and construct binary codes from them using the generalizations\n$\\phi$ and $\\Phi$ of the Gray map, then the weight enumerators of the binary\ncodes obtained will satisfy the MacWilliams identity. The classes of\n$Z_{2^k}$-linear Hadamard codes and co-$Z_{2^k}$-linear extended 1-perfect\ncodes are described, where co-$Z_{2^k}$-linearity means that the code can be\nobtained from a linear $Z_{2^k}$-code with the help of the new generalized Gray\nmap. Keywords: Gray map, Hadamard codes, MacWilliams identity, perfect codes,\n$Z_{2^k}$-linearity\n", " Previous work on convexity of neural codes has produced codes that are\nopen-convex but not closed-convex -- or vice-versa. However, why a code is one\nbut not the other, and how to detect such discrepancies are open questions. We\ntackle these questions in two ways. First, we investigate the concept of\ndegeneracy introduced by Cruz et al., and extend their results to show that\nnondegeneracy precisely captures the situation when taking closures or\ninteriors of open or closed realizations, respectively, yields another\nrealization of the code. Second, we give the first general criteria for\nprecluding a code from being closed-convex (without ruling out open-convexity),\nunifying ad-hoc geometric arguments in prior works. One criterion is built on a\nphenomenon we call a rigid structure, while the other can be stated\nalgebraically, in terms of the neural ideal of the code. These results\ncomplement existing criteria having the opposite purpose: precluding\nopen-convexity but not closed-convexity. Finally, we show that a family of\ncodes shown by Jeffs to be not open-convex is in fact closed-convex and\nrealizable in dimension two.\n", " Arguments are provided which show that extension of renormalizability in\nquantum field theory is possible. A dressed scheme for the perturbation\nexpansion is proposed. It is proven that in this scheme a nonrenormalizable\ninteraction becomes renormalizable in the restrictive sense, i.e. its\nultraviolet divergences can be cancelled by a finite number of counterterms\nincluded in the Lagrangian. As an illustration, the renormalization of the\n$\\pi$-nucleon pseudovector interaction is discussed in some detail.\n", " The chapter provides an introduction to the basic concepts of Algebraic\nTopology with an emphasis on motivation from applications in the physical\nsciences. It finishes with a brief review of computational work in algebraic\ntopology, including persistent homology.\n", " In this paper we study the main characteristics of some evaluation codes\nparameterized by the edges of a bipartite graph with a perfect matching.\n", " Arithmetic combinatorics is often concerned with the problem of bounding the\nbehaviour of arbitrary finite sets in a group or ring with respect to\narithmetic operations such as addition or multiplication. Similarly,\ncombinatorial geometry is often concerned with the problem of bounding the\nbehaviour of arbitrary finite collections of geometric objects such as points,\nlines, or circles with respect to geometric operations such as incidence or\ndistance. Given the presence of arbitrary finite sets in these problems, the\nmethods used to attack these problems have primarily been combinatorial in\nnature. In recent years, however, many outstanding problems in these questions\nhave been solved by algebraic means (and more specifically, using tools from\nalgebraic geometry and/or algebraic topology), giving rise to an emerging set\nof techniques which is now known as the polynomial method.\n While various instances of the polynomial method have been known for decades\n(e.g. Stepanov's method, the combinatorial nullstellensatz, or Baker's\ntheorem), the general theory of this method is still in the process of\nmaturing; in particular, the limitations of the polynomial method are not well\nunderstood, and there is still considerable scope to apply deeper results from\nalgebraic geometry or algebraic topology to strengthen the method further. In\nthis survey we present several of the known applications of these methods,\nfocusing on the simplest cases to illustrate the techniques. We will assume as\nlittle prior knowledge of algebraic geometry as possible.\n", " In the present paper, we introduce the concept of harmonic Tutte polynomials\nof matroids and discuss some of their properties. In particular, we generalize\nGreene's theorem, thereby expressing harmonic weight enumerators of codes as\nevaluations of harmonic Tutte polynomials.\n", " On a Goppa code whose structure polynomial has coefficients in the symbol\nfield, the Frobenius acts. Its fixed codewords form a subcode. Deleting the\nnaturally occurred redundance, we obtain a new code. It is proved that these\nnew codes approach the Gilbert-Varshamov bound. It is also proved that these\ncodes can be decoded within $O(n^2(\\logn)^a)$ operations in the symbol field,\nwhich is usually much small than the location field, where $n$ is the codeword\nlength, and $a$ a constant determined by the polynomial factorization\nalgorithm.\n", " In this paper, we introduce the concept of the complete cycle index and\ndiscuss a relation with the complete weight enumerator in coding theory. This\nwork was motivated by Cameron's lecture note \"Polynomial aspects of codes,\nmatroids and permutation groups.\"\n", " For planar polynomials systems the existence of an invariant algebraic curve\nlimits the number of limit cycles not contained in this curve. We present a\ngeneral approach to prove non existence of periodic orbits not contained in\nthis given algebraic curve. When the method is applied to parametric families\nof polynomial systems that have limit cycles for some values of the parameters,\nour result leads to effective algebraic conditions on the parameters that force\nnon existence of the periodic orbits. As applications we consider several\nfamilies of quadratic systems: the ones having some quadratic invariant\nalgebraic curve, the known ones having an algebraic limit cycle, a family\nhaving a cubic invariant algebraic curve and other ones. For any quadratic\nsystem with two invariant algebraic curves we prove a finiteness result for its\nnumber of limit cycles that only depends on the degrees of these curves. We\nalso consider some families of cubic systems having either a quadratic or a\ncubic invariant algebraic curve and a family of Lienard systems. We also give a\nnew and simple proof of the known fact that quadratic systems with an invariant\nparabola have at most one limit cycle. In fact, what we show is that this\nresult is a consequence of the similar result for quadratic systems with an\ninvariant straight line.\n", " A systematic and comprehensive study of p-adic refinement equations and\nsubdivision scheme associated with a finitely supported refinement mask are\ncarried out in this paper. The Lq -convergence of the subdivision scheme is\ncharacterized in terms of the q-norm joint spectral radii of a collection of\noperators associated with the refinement mask. Also, the smoothness of\ncomplex-valued functions on Qp is investigated.\n", " We study questions in incidence geometry where the precise position of points\nis `blurry' (e.g. due to noise, inaccuracy or error). Thus lines are replaced\nby narrow tubes, and more generally affine subspaces are replaced by their\nsmall neighborhood. We show that the presence of a sufficiently large number of\napproximately collinear triples in a set of points in d dimensional complex\nspace implies that the points are close to a low dimensional affine subspace.\nThis can be viewed as a stable variant of the Sylvester-Gallai theorem and its\nextensions.\n Building on the recently found connection between Sylvester-Gallai type\ntheorems and complex Locally Correctable Codes (LCCs), we define the new notion\nof stable LCCs, in which the (local) correction procedure can also handle small\nperturbations in the euclidean metric. We prove that such stable codes with\nconstant query complexity do not exist. No impossibility results were known in\nany such local setting for more than 2 queries.\n", " In our recent paper we characterized the extension property for symmetrized\nweight composition for linear codes over a module alphabet. Several\nimprovements for the case of vector space alphabets are given in this paper. A\ndetailed description of the property of $G$-pseudo-injectivity for vector\nspaces is made.\n", " In order to facilitate the production of codes using WiSARD-based models,\nLabZero developed an ML library C++/Python called wisardpkg. This library is an\nMIT-licensed open-source package hosted on GitHub under the license.\n", " Accessing fluid infiltration in nanogranular coatings is an outstanding\nchallenge, of relevance for applications ranging from nanomedicine to\ncatalysis. A sensing platform, allowing to quantify the amount of fluid\ninfiltrated in a nanogranular ultrathin coating, with thickness in the 10 to 40\nnm range, is here proposed and theoretically investigated by multiscale\nmodelling. The scheme relies on impulsive photoacoustic excitation of\nhypersonic mechanical breathing modes in engineered gas-phase synthesised\nnanogranular metallic ultathin films and time-resolved acousto-optical read-out\nof the breathing modes frequency shift upon liquid infiltration. A superior\nsensitivity, exceeding 26x103 cm^2/g, is predicted upon equivalent areal mass\nloading of a few ng/mm^2. The capability of the present scheme to discriminate\namong different infiltration patterns is discussed. The platform is an ideal\ntool to investigate nano fluidics in granular materials and naturally serves as\na distributed nanogetter coating, integrating fluid sensing capabilities. The\nproposed scheme is readily extendable to other nanoscale and mesoscale porous\nmaterials.\n", " Let $n$ be a positive integer and $I$ a $k$-subset of integers in $[0,n-1]$.\nGiven a $k$-tuple $A=(\\alpha_0, \\cdots, \\alpha_{k-1})\\in \\mathbb{F}^k_{q^n}$,\nlet $M_{A,I}$ denote the matrix $(\\alpha_i^{q^j})$ with $0\\leq i\\leq k-1$ and\n$j\\in I$. When $I=\\{0,1,\\cdots, k-1\\}$, $M_{A,I}$ is called a Moore matrix\nwhich was introduced by E. H. Moore in 1896. It is well known that the\ndeterminant of a Moore matrix equals $0$ if and only if $\\alpha_0,\\cdots,\n\\alpha_{k-1}$ are $\\mathbb{F}_q$-linearly dependent. We call $I$ that satisfies\nthis property a Moore exponent set. In fact, Moore exponent sets are equivalent\nto maximum rank-distance (MRD) code with maximum left and right idealisers over\nfinite fields. It is already known that $I=\\{0,\\cdots, k-1\\}$ is not the unique\nMoore exponent set, for instance, (generalized) Delsarte-Gabidulin codes and\nthe MRD codes recently discovered by Csajb\\'ok, Marino, Polverino and the\nsecond author both give rise to new Moore exponent sets. By using algebraic\ngeometry approach, we obtain an asymptotic classification result: for $q>5$, if\n$I$ is not an arithmetic progression, then there exist an integer $N$ depending\non $I$ such that $I$ is not a Moore exponent set provided that $n>N$.\n", " In my masters thesis I prove a square root bound on the distance of\nhomological codes that come from two dimensional surfaces, as a result of the\nsystolic inequality. I also give a detailed version of M.H. Freedman's proof\nthat due to systolic freedom, this bound does not hold in higher dimensions.\n", " Implicit functions provide a fundamental basis to model 3D objects, no matter\nthey are rigid or deformable, in computer graphics and geometric modeling. This\npaper introduces a new constructive scheme of implicitly-defined 3D objects\nbased on products of implicit functions. This scheme is in contrast with\npopular approaches like blobbies, meta balls and soft objects, which rely on\nthe sum of specific implicit functions to fit a 3D object to a set of spheres.\n", " We consider Cauchy type integrals $I(t)={1\\over 2\\pi i}\\int_{\\gamma}\n{g(z)dz\\over z-t}$ with $g(z)$ an algebraic function. The main goal is to give\nconstructive (at least, in principle) conditions for $I(t)$ to be an algebraic\nfunction, a rational function, and ultimately an identical zero near infinity.\nThis is done by relating the Monodromy group of the algebraic function $g$, the\ngeometry of the integration curve $\\gamma$, and the analytic properties of the\nCauchy type integrals. The motivation for the study of these conditions is\nprovided by the fact that certain Cauchy type integrals of algebraic functions\nappear in the infinitesimal versions of two classical open questions in\nAnalytic Theory of Differential Equations: the Poincar\\'e Center-Focus problem\nand the second part of the Hilbert 16-th problem.\n", " We present a method that detects boundaries of parts in 3D shapes represented\nas point clouds. Our method is based on a graph convolutional network\narchitecture that outputs a probability for a point to lie in an area that\nseparates two or more parts in a 3D shape. Our boundary detector is quite\ngeneric: it can be trained to localize boundaries of semantic parts or\ngeometric primitives commonly used in 3D modeling. Our experiments demonstrate\nthat our method can extract more accurate boundaries that are closer to\nground-truth ones compared to alternatives. We also demonstrate an application\nof our network to fine-grained semantic shape segmentation, where we also show\nimprovements in terms of part labeling performance.\n", " Many of the existing Person Re-identification (Re-ID) approaches depend on\nfeature maps which are either partitioned to localize parts of a person or\nreduced to create a global representation. While part localization has shown\nsignificant success, it uses either na{\\i}ve position-based partitions or\nstatic feature templates. These, however, hypothesize the pre-existence of the\nparts in a given image or their positions, ignoring the input image-specific\ninformation which limits their usability in challenging scenarios such as Re-ID\nwith partial occlusions and partial probe images. In this paper, we introduce a\nspatial attention-based Dynamic Part Template Initialization module that\ndynamically generates part-templates using mid-level semantic features at the\nearlier layers of the backbone. Following a self-attention layer, human\npart-level features of the backbone are used to extract the templates of\ndiverse human body parts using a simplified cross-attention scheme which will\nthen be used to identify and collate representations of various human parts\nfrom semantically rich features, increasing the discriminative ability of the\nentire model. We further explore adaptive weighting of part descriptors to\nquantify the absence or occlusion of local attributes and suppress the\ncontribution of the corresponding part descriptors to the matching criteria.\nExtensive experiments on holistic, occluded, and partial Re-ID task benchmarks\ndemonstrate that our proposed architecture is able to achieve competitive\nperformance. Codes will be included in the supplementary material and will be\nmade publicly available.\n", " In the paper we study a special parameter containing algebraic inequality\ninvolving sum of reciprocals and product of positive real numbers whose sum is\n1. We determine the best values of the parameter using a new optimization\nargument. In the case of three numbers the algebraic inequality have some\ninteresting geometric applications involving a generalization of Euler's\ninequality about the ratio of radii of circumscribed and inscribed circles of a\ntriangle.\n" ]

[ " In this paper, we introduce and study the concept of L-Dunford-Pettis sets\nand L-Dunford-Pettis property in Banach spaces. Next, we give a\ncharacterization of the L-Dunford-Pettis property with respect to some\nwell-known geometric properties of Banach spaces. Finally, some\ncomplementability of operators on Banach spaces with the L-Dunford-Pettis\nproperty are also investigated\n", " It is known that Dobrushin's ergodicity coefficient is one of the effective\ntools in the investigations of limiting behavior of Markov processes. Several\ninteresting properties of the ergodicity coefficient of a positive mapping\ndefined on ordered Banach space with a base have been studied. In this paper,\nwe consider uniformly mean ergodic and asymptotically stable Markov operators\non ordered Banach spaces. In terms of the ergodicity coefficient, we prove\nuniform mean ergodicity criterion in terms of the ergodicity coefficient.\nMoreover, we develop the perturbation theory for uniformly asymptotically\nstable Markov chains on ordered Banach spaces. In particularly, main results\nopen new perspectives in the perturbation theory for quantum Markov processes\ndefined on von Neumann algebras. Moreover, by varying the Banach spaces one can\nobtain several interesting results in both classical and quantum settings as\nwell.\n", " The first part of this article is a brief survey of the properties of\nso-called almost interior points in ordered Banach spaces. Those vectors can be\nseen as a generalization of ``functions which are strictly positive almost\neverywhere'' on $L^p$-spaces and of ``quasi-interior points'' in Banach\nlattices.\n In the second part we study the long--term behaviour of strongly positive\noperator semigroups on ordered Banach spaces; these are semigroups which, in a\nsense, map every non-zero positive vector to an almost interior point. Using\nthe Jacobs--de Leeuw--Glicksberg decomposition together with the theory\npresented in the first part of the paper we deduce sufficiency criteria for\nsuch semigroups to converge (strongly or in operator norm) as time tends to\ninfinity. This generalises known results for semigroups on Banach lattices as\nwell as on normally ordered Banach spaces with unit.\n", " This paper studies the bounded approximation property (BAP) in quasi Banach\nspaces. In the first part of the paper we show that the kernel of any\nsurjective operator $\\ell_p\\to X$ has the BAP when $X$ has it and $0<p\\leq 1$,\nwhich is an analogue of the corresponding result of Lusky for Banach spaces. We\nthen obtain and study nonlocally convex versions of the Kadec-Pe\\l\nczy\\'nski-Wojtaszczyk complementably universal spaces for Banach spaces with\nthe BAP.\n", " The main purpose of this paper is to treat semigroups properties, like norm\ncontinuity, compactness and differentiability for perturbed semigroups in\nBanach spaces. In particular, we investigate three large classes of\nperturbations, Miyadera-Voigt, Desch-Schappacher and Staffans-Weiss\nperturbations. Our approach is mainly based on feedback theory of Salamon-Weiss\nsystems. Our results are applied to abstract boundary integro-differential\nequations in Banach spaces.\n", " In this paper we consider perturbation of X_d-Bessel sequences, X_d-frames,\nBanach frames, atomic decompositions and X_d-Riesz bases in separable Banach\nspaces. Equivalence between some perturbation conditions is investigated.\n", " Studying the function spaces defined by neural networks helps to understand\nthe corresponding learning models and their inductive bias. While in some\nlimits neural networks correspond to function spaces that are reproducing\nkernel Hilbert spaces, these regimes do not capture the properties of the\nnetworks used in practice. In contrast, in this paper we show that deep neural\nnetworks define suitable reproducing kernel Banach spaces.\n These spaces are equipped with norms that enforce a form of sparsity,\nenabling them to adapt to potential latent structures within the input data and\ntheir representations. In particular, leveraging the theory of reproducing\nkernel Banach spaces, combined with variational results, we derive representer\ntheorems that justify the finite architectures commonly employed in\napplications. Our study extends analogous results for shallow networks and can\nbe seen as a step towards considering more practically plausible neural\narchitectures.\n", " In this paper, the problems of perturbation and expression for the\nMoore--Penrose metric generalized inverses of bounded linear operators on\nBanach spaces are further studied. By means of certain geometric assumptions of\nBanach spaces, we first give some equivalent conditions for the Moore--Penrose\nmetric generalized inverse of perturbed operator to have the simplest\nexpression $T^M(I+ \\delta TT^M)^{-1}$. Then, as an application our results, we\ninvestigate the stability of some operator equations in Banach spaces under\ndifferent type perturbations.\n", " In 1992, Kiendi, Adamy and Stelzner investigated under which conditions a\ncertain type of function constituted a Lyapunov function for some\ntime-invariant linear system. Six years later, it was obtained that this\nproperty holds if and only if the Banach space enjoys the self-extension\nproperty. However, the knowledge of these spaces needed to be extended in order\nto make useful this characterization, since there were little information on\nwhich classic Banach spaces satisfy this property and its relations with other\nclassic properties. We present the self-extension property in a wider frame,\nrelating it to other well-known spaces as the $1$-injective or $1$-projective\nones. We investigate the property in two important low-dimensional classic\nBanach spaces: $\\mathbb{R}_1^3$ and $\\mathbb{R}_1^4$. We introduce the concept\nof $k-$self-extensible spaces and a discussion of the stability of the\nproperty. We also show that every real Banach space of dimension greater than\nor equal to $3$ can be equivalently renormed to fail the self-extension\nproperty. Finally, we summarize some consequences of our study and mention some\nopen questions which appear naturally.\n", " We shall prove a rearrangement inequality in probability measure spaces in\norder to obtain sharp Leibniz-type rules of mean oscillations in Lp-spaces and\nrearrangement invariant Banach function spaces.\n", " In this paper, we study the descriptive complexity of some inevitable classes\nof Banach spaces. Precisely, as shown in [Go], every Banach space either\ncontains a hereditarily indecomposable subspace or an unconditional basis, and,\nas shown in [FR], every Banach space either contains a minimal subspace or a\ncontinuously tight subspace. In these notes, we study the complexity of those\ninevitable classes as well as the complexity of containing a subspace in any of\nthose classes.\n", " We present condition on higher order asymptotic behaviour of basic sequences\nin a Banach space ensuring the existence of bounded non-compact strictly\nsingular operator on a subspace. We apply it in asymptotic $\\ell_p$ spaces,\n$1\\leq p<\\infty$, in particular in convexified mixed Tsirelson spaces and\nrelated asymptotic $\\ell_p$ HI spaces.\n", " We recall the Chernoff-Marsden definition of weak symplectic structure and\ngive a rigorous treatment of the functional analysis and geometry of weak\nsymplectic Banach spaces. We define the Maslov index of a continuous path of\nFredholm pairs of Lagrangian subspaces in continuously varying Banach spaces.\nWe derive basic properties of this Maslov index and emphasize the new features\nappearing.\n", " Let $f$ be a distribution (generalised function) on the real line. If there\nis a continuous function $F$ with real limits at infinity such that $F'=f$\n(distributional derivative) then the distributional integral of $f$ is defined\nas $\\int_{-\\infty}^\\infty f = F(\\infty) - F(-\\infty)$. It is shown that this\nsimple definition gives an integral that includes the Lebesgue and\nHenstock--Kurzweil integrals. The Alexiewicz norm leads to a Banach space of\nintegrable distributions that is isometrically isomorphic to the space of\ncontinuous functions on the extended real line with uniform norm. The dual\nspace is identified with the functions of bounded variation. Basic properties\nof integrals are established using elementary properties of distributions:\nintegration by parts, H\\\"older inequality, change of variables, convergence\ntheorems, Banach lattice structure, Hake theorem, Taylor theorem, second mean\nvalue theorem. Applications are made to the half plane Poisson integral and\nLaplace transform. The paper includes a short history of Denjoy's descriptive\nintegral definitions. Distributional integrals in Euclidean spaces are\ndiscussed and a more general distributional integral that also integrates Radon\nmeasures is proposed.\n", " In this paper, we establish a Minkowski-type inequality for weak Lebesgue\nspace, which allows us to obtain a characterization of relative compactness in\nthese spaces. Furthermore, we are the first to investigate the compactness\nresults of commutators at the endpoint. The paper provides a comprehensive\nstudy of the compactness properties of commutators of Calder\\'{o}n-Zygmund\noperators in Hardy and $L^{1}(\\mathbb{R}^n)$ type spaces. Additionally, we\nprovide factorization theorems for Hardy spaces in terms of singular integral\noperators in the $L^1(\\mathbb{R}^n)$ space.\n", " Let $(e_i)$ be a fundamental system of a Banach space.\n We consider the problem of approximating linear combinations of elements of\nthis system by linear combinations using quantized coefficients. We will\nconcentrate on systems which are possibly redundant. Our model for this\nsituation will be frames in Banach spaces.\n", " We create a new, functional calculus, approach to approximation of\nC_0-semigroups on Banach spaces. As an application of this approach, we obtain\noptimal convergence rates in classical approximation formulas for\nC_0-semigroups. In fact, our methods allow one to derive a number of similar\nformulas and equip them with sharp convergence rates. As a byproduct, we prove\na new interpolation principle leading to efficient norm estimates in the Banach\nalgebra of Laplace transforms of bounded measures on the semi-axis.\n", " A topological space is said to be sequential if every sequentially closed\nsubspace is closed. We consider Banach spaces with weak*-sequential dual ball.\nIn particular, we show that if $X$ is a Banach space with weak*-sequentially\ncompact dual ball and $Y \\subset X$ is a subspace such that $Y$ and $X/Y$ have\nweak*-sequential dual ball, then $X$ has weak*-sequential dual ball. As an\napplication we obtain that the Johnson-Lindenstrauss space $JL_2$ and $C(K)$\nfor $K$ scattered compact space of countable height are examples of Banach\nspaces with weak*-sequential dual ball, answering in this way a question of A.\nPlichko.\n", " In 1960, Klee showed that a subset of a Euclidean space must be a singleton\nprovided that each point in the space has a unique farthest point in the set.\nThis classical result has received much attention; in fact, the Hilbert space\nversion is a famous open problem. In this paper, we consider Klee sets from a\nnew perspective. Rather than measuring distance induced by a norm, we focus on\nthe case when distance is meant in the sense of Bregman, i.e., induced by a\nconvex function. When the convex function has sufficiently nice properties,\nthen - analogously to the Euclidean distance case - every Klee set must be a\nsingleton. We provide two proofs of this result, based on Monotone Operator\nTheory and on Nonsmooth Analysis. The latter approach leads to results that\ncomplement work by Hiriart-Urruty on the Euclidean case.\n", " We show a Wolff-Denjoy type theorem in complete geodesic spaces in the spirit\nof Beardon's framework that unifies several results in this area. In\nparticular, it applies to strictly convex bounded domains in $\\mathbb{R}^{n}$\nor $\\mathbb{C}^{n}$ with respect to a large class of metrics including\nHilbert's and Kobayashi's metrics. The results are generalized to $1$-Lipschitz\ncompact mappings in infinite-dimesional Banach spaces.\n", " We prove that no ultraproduct of Banach spaces via a countably incomplete\nultrafilter can contain $c_0$ complemented. This shows that a \"result\" widely\nused in the theory of ultraproducts is wrong. We then amend a number of results\nwhose proofs had been infected by that statement. In particular we provide\nproofs for the following statements: (i) All $M$-spaces, in particular all\n$C(K)$-spaces, have ultrapowers isomorphic to ultrapowers of $c_0$, as well as\nall their complemented subspaces isomorphic to their square. (ii) No ultrapower\nof the Gurari\\u \\i\\ space can be complemented in any $M$-space. (iii) There\nexist Banach spaces not complemented in any $C(K)$-space having ultrapowers\nisomorphic to a $C(K)$-space.\n", " In this work we investigate the background dynamics when dark energy is\ncoupled to dark matter with a suitable interaction in the universe described by\nLoop quantum cosmology. Dark energy in the form of Generalised Cosmic Chaplygin\ngas is considered. A suitable interaction between dark energy and dark matter\nis taken into account in order to at least alleviate (if not solve) the cosmic\ncoincidence problem. The dynamical system of equations is solved numerically\nand a stable scaling solution is obtained. A significant attempt towards the\nsolution of the cosmic coincidence problem is taken. The statefinder parameters\nare also calculated to classify the dark energy model. Graphs and phase\ndiagrams are drawn to study the variations of these parameters. It is seen that\nthe background dynamics of Generalised Cosmic Chaplygin gas is completely\nconsistent with the notion of an accelerated expansion in the late universe.\nFrom the graphs, generalised cosmic Chaplygin gas is identified as a dark fluid\nwith a lesser negative pressure compared to Modified Chaplygin gas, thus\nsupporting a 'No Big Rip' cosmology. It has also been shown that in this model\nthe universe follows the power law form of expansion around the critical point,\nwhich is consistent with the known results. Future singularities that may be\nformed in this model as an ultimate fate of the universe has been studied in\ndetail. It was found that the model is completely free from any types of future\nsingularities.\n", " Starting from the meaning of the conjugate of a complex Hilbert space,\nincluding a related application of the theorem of Fr\\'{e}chet-Riesz (by which\nan analysis of semilinear operators can be reduced to - linear - operator\ntheory) to a revisit of applications of nuclear and absolutely $p$-summing\noperators in algebraic quantum field theory in the sense of Araki, Haag and\nKastler ($p=2$) and more recently in the framework of general probabilistic\nspaces ($p=1$), we will outline that Banach operator ideals in the sense of\nPietsch, or equivalently tensor products of Banach spaces in the sense of\nGrothendieck are even lurking in the foundations and philosophy of quantum\nphysics and quantum information theory. In particular, we concentrate on their\nimportance in algebraic quantum field theory. In doing so, we establish a\ncanonical isometric isomorphism between the Hilbert spaces $H\\otimes_2 (K\n\\otimes_2 L)$ and $(H \\otimes_2 K) \\otimes_2 L$ (Theorem 3.8) and revisit the\nrole of trace class operators.\n A few applications are specified, including the appropriateness of the class\nof Hilbert-Schmidt operators and an implied Banach operator ideal\nrepresentation of the tensor product of two complex Hilbert spaces $H \\otimes_2\nK$ (Proposition 3.4) and a purely linear algebraic description of the quantum\nteleportation process (Example 3.10).\n", " A detailed theory of stochastic integration in UMD Banach spaces has been\ndeveloped recently by the authors. The present paper is aimed at giving various\nsufficient conditions for stochastic integrability.\n", " We study the general measures of non-compactness defined on subsets of a dual\nBanach space, their associated derivations and their $\\omega$-iterates. We\nintroduce the notions of convexifiable and sublinear measure of non-compactness\nand investigate the properties of its associated fragment and slice\nderivations. We apply our results to the Kuratowski measure of non-compactness\nand to the study of the Szlenk index of a Banach space. As a consequence, we\nobtain that the Szlenk index and the convex Szlenk index of a separable Banach\nspace are always equal. We also give, for any countable ordinal $\\alpha$, a\ncharacterization of the Banach spaces with Szlenk index bounded by\n$\\omega^{\\alpha+1}$ in terms of the existence of an equivalent renorming. This\nextends a result by Knaust, Odell and Schlumprecht on Banach spaces with Szlenk\nindex equal to $\\omega$.\n", " We investigate the log-concavity on the half-line of the Wright function\n$\\phi(-\\alpha,\\beta,-x),$ in the probabilistic setting $\\alpha\\in (0,1)$ and\n$\\beta \\ge 0.$ Applications are given to the construction of generalized\nentropies associated to the corresponding Mittag-Leffler function. A natural\nconjecture for the equivalence between the log-concavity of the Wright function\nand the existence of such generalized entropies is formulated. The problem is\nsolved for $\\beta\\geq\\alpha$ and in the classical case $\\beta = 1-\\alpha$ of\nthe Mittag-Leffler distribution, which exhibits a certain critical parameter\n$\\alpha_*= 0.771667...$ defined implicitly on the Gamma function and\ncharacterizing the log-concavity. We also prove that the probabilistic Wright\nfunctions are always unimodal, and that they are multiplicatively strongly\nunimodal if and only if $\\beta\\geq\\alpha$ or $\\alpha\\le 1/2$ and $\\beta = 0.$\n", " In this paper, we present some common fixed point theorems for a commuting\npair of mappings, including a generalized nonexpansive single valued mapping\nand a generalized nonexpansive multivalued mapping in strictly convex Banach\nspaces. The results obtained in this paper extend and improve some recent known\nresults.\n", " We study properties of representing and absolutely representing systems of\nsubspaces in Banach spaces. We also present sufficient conditions for the\nsystem of subspaces to be a representing system of subspaces.\n", " We obtain an equivalent implicit characterization of $L^p$ Banach spaces that\nis amenable to a logical treatment. Using that, we obtain an axiomatization for\nsuch spaces into a higher-order logical system, the kind of which is used in\nproof mining, a research program that aims to obtain the hidden computational\ncontent of mathematical proofs using tools from mathematical logic. As an\naside, we obtain a concrete way of formalizing $L^p$ spaces in positive-bounded\nlogic. The axiomatization is followed by a corresponding metatheorem in the\nstyle of proof mining. We illustrate its use with the derivation for this class\nof spaces of the standard modulus of uniform convexity.\n", " Let $A$ be a right linear operator on a two-sided quaternionic Banach space\n$X$. The paper studies the Drazin inverse for right linear operators on a\nquaternionic Banach space. It is shown that if $A$ is Drazin invertible then\nthe Drazin inverse of $A$ is given by $f(A)$ where $f$ is $0$ in an axially\nsymmetric neighborhood of $0$ and $f(q) = q^{-1}$ in an axially symmetric\nneighborhood of the nonzero spherical spectrum of $A$. Some results analogous\nto the ones concerning the Drazin inverse of operators on complex Banach spaces\nare proved in the quaternionic context.\n", " We introduce two Bishop-Phelps-Bollob\\'as moduli which measure, for a given\nBanach space, what is the best possible Bishop-Phelps-Bollob\\'as theorem in\nthis space. We show that there is a common upper bound for these moduli for all\nBanach spaces and we present an example showing that this bound is sharp. We\nprove the continuity of these moduli and an inequality with respect to duality.\nWe calculate the two moduli for Hilbert spaces and also present many examples\nfor which the moduli have the maximum possible value (among them, there are\n$C(K)$ spaces and $L_1(\\mu)$ spaces). Finally, we show that if a Banach space\nhas the maximum possible value of any of the moduli, then it contains almost\nisometric copies of the real space $\\ell_\\infty^{(2)}$ and present an example\nshowing that this condition is not sufficient.\n", " We establish estimates for the Lebesgue parameters of the Chebyshev Weak\nThresholding Greedy Algorithm in the case of general bases in Banach spaces.\nThese generalize and slightly improve earlier results in [9], and are\ncomplemented with examples showing the optimality of the bounds. Our results\nalso correct certain bounds recently announced in [18], and answer some\nquestions left open in that paper.\n", " We introduce the class of weighted \"rotation-like\" operators and study\ngeneral properties of essential spectra of such operators. Then we use this\napproach to investigate and in some cases completely describe essential spectra\nof weighted rotation operators in Banach spaces of measurable and analytic\nfunctions.\n", " This paper reviews the main estimation and prediction results derived in the\ncontext of functional time series, when Hilbert and Banach spaces are\nconsidered, specially, in the context of autoregressive processes of order one\n(ARH(1) and ARB(1) processes, for H and B being a Hilbert and Banach space,\nrespectively). Particularly, we pay attention to the estimation and prediction\nresults, and statistical tests, derived in both parametric and non-parametric\nframeworks. A comparative study between different ARH(1) prediction approaches\nis developed in the simulation study undertaken.\n", " In this paper, we give some sufficient and necessary conditions for cosine\noperator functions on solid Banach function spaces to be chaotic or\ntopologically transitive.\n", " Jacobi-Trudy formula for a generalisation of Schur polynomials related to any\nsequence of orthogonal polynomials in one variable is given. As a corollary we\nhave Giambelli formula for generalised Schur polynomials.\n", " We provide a new, short proof of the density in energy of Lipschitz functions\ninto the metric Sobolev space defined by using plans with barycenter (and thus,\na fortiori, into the Newtonian-Sobolev space). Our result covers first-order\nSobolev spaces of exponent $p\\in(1,\\infty)$, defined over a complete and\nseparable metric space endowed with a boundedly-finite Borel measure.\n Our proof is based on a completely smooth analysis: first we reduce the\nproblem to the Banach space setting, where we consider smooth functions instead\nof Lipschitz ones, then we rely on classical tools in convex analysis and on\nthe superposition principle for normal $1$-currents. Along the way, we obtain a\nnew proof of the density in energy of smooth cylindrical functions in Sobolev\nspaces defined over a separable Banach space endowed with a finite Borel\nmeasure.\n", " We study certain interpolation and extension properties of the space of\nregular operators between two Banach lattices. Let $R_p$ be the space of all\nthe regular (or equivalently order bounded) operators on $L_p$ equipped with\nthe regular norm. We prove the isometric identity $R_p = (R_\\infty,R_1)^\\theta$\nif $\\theta = 1/p$, which shows that the spaces $(R_p)$ form an interpolation\nscale relative to Calder\\'on's interpolation method. We also prove that if\n$S\\subset L_p$ is a subspace, every regular operator $u : S \\to L_p$ admits a\nregular extension $\\tilde u : L_p \\to L_p$ with the same regular norm. This\nextends a result due to Mireille L\\'evy in the case $p = 1$. Finally, we apply\nthese ideas to the Hardy space $H^p$ viewed as a subspace of $L_p$ on the\ncircle. We show that the space of regular operators from $H^p$ to $L_p$\npossesses a similar interpolation property as the spaces $R_p$ defined above.\n", " In the setting of Banach lattices the weak (resp. positive) Grothendieck\nspaces have been defined. We localize such notions by defining new classes of\nsets that we study and compare with some quite related different classes. This\nallows us to introduce and compare the corresponding linear operators.\n", " Operator representation of Cole-Hopf transform is obtained based on the\nlogarithmic representation of infinitesimal generators. For this purpose the\nrelativistic formulation of abstract evolution equation is introduced. Even\nindependent of the spatial dimension, the Cole-Hopf transform is generalized to\na transform between linear and nonlinear equations defined in Banach spaces. In\nconclusion a role of transform between the evolution operator and its\ninfinitesimal generator is understood in the context of generating nonlinear\nsemigroup.\n", " We study the Banach spectral gap lambda_1(G;X,p) of finite graphs G for pairs\n(X,p) of Banach spaces and exponents. We define the notion of sphere\nequivalence between Banach spaces and show a generalization of Matousek's\nextrapolation for Banach spaces sphere equivalent to uniformly convex ones. As\na byproduct, we prove that expanders are automatically expanders with respects\nto (X,p) for any X sphere equivalent to a uniformly curved Banach space and for\nany p strictly bigger than 1.\n", " Several local geometric properties of Orlicz space $L_\\phi$ are presented for\nan increasing Orlicz function $\\phi$ which is not necessarily convex, and thus\n$L_\\phi$ does not need to be a Banach space. In addition to monotonicity of\n$\\phi$ it is supposed that $\\phi(u^{1/p})$ is convex for some $p>0$ which is\nequivalent to that its lower Matuszewska-Orlicz index $\\alpha_\\phi>0$. Such\nspaces are locally bounded and are equipped with natural quasi-norms. Therefore\nmany local geometric properties typical for Banach spaces can also be studied\nin those spaces. The techniques however have to be different, since duality\ntheory cannot be applied in this case. In this article we present complete\ncriteria, in terms of growth conditions of $\\phi$, for $L_\\phi$ to have type\n$0<p\\le2$, cotype $q\\ge 2$, to be (order) $p$-convex or $q$-concave, to have an\nupper $p$-estimate or a lower $q$-estimate, for $0<p,q<\\infty$. We provide\ndetailed proofs of most results, avoiding appealing to general not necessary\ntheorems.\n", " We prove an analogue of Bonami's (hypercontractive) lemma for complex-valued\nfunctions on $\\mathcal{L}(V,W)$, where $V$ and $W$ are vector spaces over a\nfinite field. This inequality is useful for functions on $\\mathcal{L}(V,W)$\nwhose `generalised influences' are small, in an appropriate sense. It leads to\na significant shortening of the proof of a recent seminal result by Khot,\nMinzer and Safra that pseudorandom sets in Grassmann graphs have near-perfect\nexpansion, which (in combination with the work of Dinur, Khot, Kindler, Minzer\nand Safra) implies the 2-2 Games conjecture (the variant, that is, with\nimperfect completeness).\n", " By extending the notion of Arens regularity of bilinear mappings, we say that\na bounded trilinear map on Banach spaces id Aron--Berner regular when all its\nsix Aron-Berner extensions to the bidual spaces coincide. We give some results\non the Aron-Berner regularity of certain trilinear maps. We then focus on the\nbidual, $E^{**},$ of a Jordan banach triple system $(E,\\pi)$, and investigate\nthose conditions under which $E^{**}$ is itself a Jordan Banach triple system\nunder each of the Aron-Berner extensions of the triple product $\\pi.$ We also\ncompare these six triple products with those arising from certain ultrafilters\nbased on the ultrapower formulation of the principle of local reflexivity. In\nparticular, we examine the Aron--Berner triple products on the bidual of a\nJB$^*$-triple in relation with the so-called Dineen's theorem. Some\nilluminating examples are included and some questions are also left undecided.\n", " We fix a gap in the proof of a result in our earlier paper \"Generalised\nnotions of amenability, II\" (J. Funct. Anal. 254 (2008) 1776-1810), and so\nprovide a new proof to a characterization of amenability for Beurling algebras.\nThe result answers a question raised by M. C. White (Bull. London. Math. Soc.\n23 (1991) 375-380).\n", " Continuous first-order logic is used to apply model-theoretic analysis to\nanalytic structures (e.g. Hilbert spaces, Banach spaces, probability spaces,\netc.). Classical computable model theory is used to examine the algorithmic\nstructure of mathematical objects that can be described in classical\nfirst-order logic. The present paper shows that probabilistic computation\n(sometimes called randomized computation) can play an analogous role for\nstructures described in continuous first-order logic. The main result of this\npaper is an effective completeness theorem, showing that every decidable\ncontinuous first-order theory has a probabilistically decidable model. Later\nsections give examples of the application of this framework to various classes\nof structures, and to some problems of computational complexity theory.\n", " The problem involving the extension of functions from a certain class and\ndefined on subdomains of the ambient space to the whole space is an old and a\nwell investigated theme in analysis. A related question whether the extensions\nthat result in the process may be chosen in a linear or a continuous manner\nbetween appropriate spaces of functions turns out to be highly nontrivial. That\nthis holds for the class of continuous functions defined on metric spaces is\nthe well-known Borsuk-Dugundji theorem which asserts that given a metric space\nM and a subspace S of M, each continuous function g on S can be extended to a\ncontinuous function f on X such that the resulting assignment from C(S) to C(M)\nis a norm-one continuous linear extension operator.\n The present paper is devoted to an investigation of this problem in the\ncontext of extendability of Lipschitz functions from closed subspaces of a\ngiven Banach space to the whole space such that the choice of the extended\nfunction gives rise to a bounded linear (extension) operator between\nappropriate spaces of Lipschitz functions. It is shown that the indicated\nproperty holds precisely when the underlying space is isomorphic to a Hilbert\nspace. Among certain useful consequences of this theorem, we provide an\nisomorphic analogue of a well-known theorem of S. Reich by show ing that closed\nconvex subsets of a Banach space X arise as Lipschitz retracts of X precisely\nwhen X is isomorphically a Hilbert space. We shall also discuss the issue of\nbounded linear extension operators between spaces of Lipschitz functions now\ndefined on arbitrary subsets of Banach spaces and provide a direct proof of the\nknown non-existence of such an extension operator by using methods which are\nmore accessible than those initially employed by the authors.\n", " The relation between the free Banach lattice generated by a Banach space and\nfree dual spaces is clarified. In particular, it is shown that for every Banach\nspace $E$ the free $p$-convex Banach lattice generated by $E^{**}$, denoted\n$FBL^p[E^{**}]$, admits a canonical isometric lattice embedding into\n$FBL^p[E]^{**}$ and $FBL^p[E^{**}]$ is lattice finitely representable in\n$FBL^p[E]$. Moreover, we also show that for $p>1$, $FBL^p[E]^{**}$ can actually\nbe considered as the free dual $p$-convex Banach lattice generated by $E$,\nwhereas for $p=1$ this happens precisely when $E$ does not contain complemented\ncopies of $\\ell_1$.\n", " One classical result in greedy approximation theory is that almost-greedy and\nsemi-greedy bases are equivalent in the context of Schauder bases in Banach\nspaces with finite cotype. This result was proved by S. J. Dilworth, N. J.\nKalton and D. Kutzarova in [7] and, recently, the first author in [2] proved\nthat the condition of finite cotype can be removed in this result. In [11], the\nauthors extend the notion of semi-greediness to the context of weights and\nproved the following: if w is a weight and $\\mathcal B$ is a Schauder basis in\na Banach space $\\mathbb X$ with finite cotype, then w-semi-greediness and\nw-almost-greediness are equivalent notions. In this paper, we prove the same\ncharacterization but removing the condition of finite cotype and, also, we try\nto relax the condition of Schauder in the characterization of semi-greediness\nusing the $\\rho$-admissibility, notion introduced recently in [4].\n", " We show that the non-compact generalised analytic Volterra operators $T_g$,\nwhere $g \\in \\mathit{BMOA}$, have the following structural rigidity property on\nthe Hardy spaces $H^p$ for $1 \\le p < \\infty$ and $p \\neq 2$: if $T_g$ is\nbounded below on an infinite-dimensional subspace $M \\subset H^p$, then $M$\ncontains a subspace linearly isomorphic to $\\ell^p$. This implies in particular\nthat any Volterra operator $T_g\\colon H^p \\to H^p$ is $\\ell^2$-singular for $p\n\\neq 2$.\n" ]

[ " Storage architectures ranging from minimum bandwidth regenerating encoded\ndistributed storage systems to declustered-parity RAIDs can be designed using\ndense partial Steiner systems in order to support fast reads, writes, and\nrecovery of failed storage units. In order to ensure good performance,\npopularities of the data items should be taken into account and the frequencies\nof accesses to the storage units made as uniform as possible. A proposed\ncombinatorial model ranks items by popularity and assigns data items to\nelements in a dense partial Steiner system so that the sums of ranks of the\nelements in each block are as equal as possible. By developing necessary\nconditions in terms of independent sets, we demonstrate that certain Steiner\nsystems must have a much larger difference between the largest and smallest\nblock sums than is dictated by an elementary lower bound. In contrast, we also\nshow that certain dense partial $S(t, t+1, v)$ designs can be labeled to\nrealize the elementary lower bound. Furthermore, we prove that for every\nadmissible order $v$, there is a Steiner triple system $(S(2, 3, v))$ whose\nlargest difference in block sums is within an additive constant of the lower\nbound.\n", " We propose and study a class-expansion/innovation/loss model of genome\nevolution taking into account biological roles of genes and their constituent\ndomains. In our model numbers of genes in different functional categories are\ncoupled to each other. For example, an increase in the number of metabolic\nenzymes in a genome is usually accompanied by addition of new transcription\nfactors regulating these enzymes. Such coupling can be thought of as a\nproportional \"recipe\" for genome composition of the type \"a spoonful of sugar\nfor each egg yolk\". The model jointly reproduces two known empirical laws: the\ndistribution of family sizes and the nonlinear scaling of the number of genes\nin certain functional categories (e.g. transcription factors) with genome size.\nIn addition, it allows us to derive a novel relation between the exponents\ncharacterising these two scaling laws, establishing a direct quantitative\nconnection between evolutionary and functional categories. It predicts that\nfunctional categories that grow faster-than-linearly with genome size to be\ncharacterised by flatter-than-average family size distributions. This relation\nis confirmed by our bioinformatics analysis of prokaryotic genomes. This proves\nthat the joint quantitative trends of functional and evolutionary classes can\nbe understood in terms of evolutionary growth with proportional recipes.\n", " In the paper, we investigate weighted composition operators on Bergman spaces\nof a half-plane. We characterize weighted composition operators which are\nhermitian and those which are complex symmetric with respect to a family of\nconjugations. As it turns out, weighted composition operators enhanced by a\nsymmetry must be bounded. Hermitian, and unitary weighted composition operators\nare proven to be complex symmetric with respect to an adapted and highly\nrelevant conjugation. We classify which the linear fractional functions give\nrise to the complex symmetry of bounded composition operators. We end the paper\nwith a natural link to complex symmetry in Lebesgue space.\n", " In the context of analytic functions on the open unit disk, a weighted\ncomposition operator is simply a composition operator followed by a\nmultiplication operator. The class of weighted composition operators has an\nimportant place in the theory of Banach spaces of analytic functions; for\ninstance, it includes all isometries on $H^p$ $(p\\neq 2)$. Very recently it was\nshown that only weighted composition operators preserve the class of outer\nfunctions.\n The present paper considers a particular question motivated by applications:\nWhich smallest possible sets of test functions can be used to identify an\nunknown weighted composition operator? This stems from a practical problem in\nsignal processing, where one seeks to identify an unknown minimum phase\npreserving operator on $L^2(\\real_+)$ using test signals.\n It is shown in the present paper that functions that determine weighted\ncomposition operators are directly linked to the classical normal family of\nschlicht functions. The main result is that a pair of functions $\\{f,g\\}$\ndistinguishes between any two weighted composition operators if and only if\nthere exists a zero-free function $h$ and a schlicht function $\\sigma$ such\nthat $\\spn\\{f,g\\}=\\spn\\{h\\sigma,h\\}$. This solves completely the underlying\nsignal processing problem and brings to light an intriguing geometric object,\nthe manifold of planes of the form $\\spn\\{h\\sigma,h\\}$. As an application of\nthe main result, it is proven that there exist compactly supported pairs in\n$L^2(\\real_+)$ that can be used to identify minimum phase preserving operators.\n", " Let $\\mathbb F$ be a finite field and let $\\mathcal A$ and $\\mathcal B$ be\nvector spaces of $\\mathbb F$-valued continuous functions defined on locally\ncompact spaces $X$ and $Y$, respectively. We look at the representation of\nlinear bijections $H:\\mathcal A\\longrightarrow \\mathcal B$ by continuous\nfunctions $h:Y\\longrightarrow X$ as weighted composition operators. In order to\ndo it, we extend the notion of Hamming metric to infinite spaces. Our main\nresult establishes that under some mild conditions, every Hamming isometry can\nbe represented as a weighted composition operator. Connections to coding theory\nare also highlighted.\n", " Necessary and sufficient conditions are already known in the Hardy spaces of\nboth the disc and the half plane for a composition operator to be an isometry,\nby Nordgren in the disc and by Chalendar and Partington in the half plane. All\nthe same, conditions for such an operator to be similar to an isometry have\ntaken much longer to find. We present some necessary conditions for general\nweighted composition operators to be similar to an isometry, and use them to\nproduce a complete characterisation of the rational composition on $H^p(\\C^+)$\nwhich have this property.\n", " This paper provides a class of complex symmetric weighted composition\noperators on $H^2(\\mathbb{D})$ to includes the unitary subclass, the Hermitian\nsubclass and the normal subclass obtained by Bourdon and Noor. A\ncharacterization of algebraic weighted composition operator with degree no more\nthan two is provided to illustrate that the weight function of a complex\nsymmetric weighted composition operator is not necessarily linear fractional.\n\\end{abstract}\n", " This paper studies mathematical properties of reaction systems that was\nintroduced by Enrenfeucht and Rozenberg as computational models inspired by\nbiochemical reaction in the living cells. In particular, we continue the study\non the generative power of functions specified by minimal reaction systems\nunder composition initiated by Salomaa. Allowing degenerate reaction systems,\nfunctions specified by minimal reaction systems over a quarternary alphabet\nthat are permutations generate the alternating group on the power set of the\nbackground set.\n", " In this paper, we completely characterize the order boundedness of weighted\ncomposition operators between different weighted Dirichlet spaces and different\nderivative Hardy spaces.\n", " We show that if you represent all primes with less than n-digits as points in\nn-dimensional space, then they can be stored and retrieved conveniently using\nn-dimensional geometry. Also once you have calculated all the prime numbers\nless than n digits, it is very easy to find out if a given number having less\nthan n-digits is or is not a prime. We do this by separating all the primes\nwhich are represented by points in n-dimension space by planes. It so turns out\nthat the number of planes q, required to separate all the points represented by\nprimes less than n-digit, are very few in number. Thus we obtain a very\nefficient storage and retrieval system in n-dimensional space. In addition the\nstorage and retieval repository has the property that when new primes are added\nthere is no need to start all over, we can begin where we last left off and add\nthe new primes in the repository and add new planes that separate them as and\nwhen necessary. Also we can arrange matters such that the repository can begin\nto accept larger primes which has more digits say n' where n' > n. The\nalgorithm does not make use of any property of prime numbers or of integers in\ngeneral,except for the fact that any n-digit integer can be represented as a\npoint in n-dimension space. Therefore the method can serve to be a storage and\nretrieval repository of any set of given integers, in practical cases they can\nrepresent information. Thus the algorithm can be used to devise a very\nefficient storage and retrieval system for large amounts of digital data.\n", " In this paper we lay special stress on analyzing the topological properties\nof the lattice systems and try to ovoid the conventional ways to calculate the\ncritical points. Only those clusters with finite sizes can execute the self\nsimilar transformations of infinite hierarchies. Each ordered cluster has\nfractal dimension, their minimum relates to the edge fixed point, which accords\nwith the transformations fixed point relating to a critical point. There are\ntwo classes of systems and clusters by their connectivity. Using mathematic\nmapping method we set up cluster-spin Gaussian model solved accurately. There\nare single state and -fold coupling state in a reducible cluster, each of which\ncorresponds to a subsystem described by a Gaussian model. By the minimal\nfractal dimension a final expression of the critical points is obtained. The\ncritical points of three lattice systems are calculated, our method makes the\ncalculations very simplified and highly accurate. A possibility of existence of\ndifferent clusters at the critical temperature is discussed.\n", " This paper intends to lay the theoretical foundation for the method of\nfunctional maps, first presented in 2012 by Ovsjanikov, Ben-Chen, Solomon,\nButscher and Guibas in the field of the theory and numerics of maps between\nshapes. We show how to analyze this method by looking at it as an application\nof the theories of composition operators, of matrix representa- tion of\noperators on separable Hilbert spaces, and of the theory of the Finite Section\nMethod. These are three well known fruitful topics in functional analysis. When\napplied to the task of modelling of correspondences of shapes in\nthree-dimensional space, these concepts lead directly to functional maps and\nits associated functional matrices. Mathematically spoken, functional maps are\ncomposition operators between two-dimensional manifolds, and functional\nmatrices are infinite matrix representations of such maps. We present an\nintroduction into the notion and theoretical foundation of the functional\nanalytic framework of the theory of matrix repre- sentation, especially of\ncomposition operators. We will also discuss two numerical methods for solving\nequations with such operators, namely, two variants of the Rectangular Finite\nSection Method. While one of these, which is well known, leads to an\noverdetermined system of linear equations, in the second one the minimum-norm\nsolution of an underdetermined system has to be computed. We will present the\nmain convergence results related to these methods.\n", " The celebrated S-Lemma was originally proposed to ensure the existence of a\nquadratic Lyapunov function in the Lur'e problem of absolute stability. A\nquadratic Lyapunov function is, however, nothing else than a squared Euclidean\nnorm on the state space (that is, a norm induced by an inner product). A\nnatural question arises as to whether squared non-Euclidean norms\n$V(x)=\\|x\\|^2$ may serve as Lyapunov functions in stability problems. This\npaper presents a novel non-polynomial S-Lemma that leads to constructive\ncriteria for the existence of such functions defined by weighted $\\ell_p$\nnorms. Our generalized S-Lemma leads to new absolute stability and absolute\ncontractivity criteria for Lur'e-type systems, including, for example, a new\nsimple proof of the Aizerman and Kalman conjectures for positive Lur'e systems.\n", " For a matrix A_2 weight W on R^p, we introduce a new notion of\nW-Calder\\'on-Zygmund matrix kernels, following earlier work in by Isralowitz.\nWe state and prove a T1 theorem for such operators and give a representation\ntheorem in terms of dyadic W-Haar shifts and paraproducts, in the spirit of\nHyt\\\"onen's Representation Theorem. Finally, by means of a Bellman function\nargument, we give sharp bounds for such operators in terms of bounds for\nweighted matrix martingale transforms and paraproducts.\n", " In hierarchical planning for Markov decision processes (MDPs), temporal\nabstraction allows planning with macro-actions that take place at different\ntime scale in form of sequential composition. In this paper, we propose a novel\napproach to compositional reasoning and hierarchical planning for MDPs under\ntemporal logic constraints. In addition to sequential composition, we introduce\na composition of policies based on generalized logic composition: Given\nsub-policies for sub-tasks and a new task expressed as logic compositions of\nsubtasks, a semi-optimal policy, which is optimal in planning with only\nsub-policies, can be obtained by simply composing sub-polices. Thus, a\nsynthesis algorithm is developed to compute optimal policies efficiently by\nplanning with primitive actions, policies for sub-tasks, and the compositions\nof sub-policies, for maximizing the probability of satisfying temporal logic\nspecifications. We demonstrate the correctness and efficiency of the proposed\nmethod in stochastic planning examples with a single agent and multiple task\nspecifications.\n", " Linear carbon chains (LCCs) are the ultimate 1D molecular system and they\nshow unique mechanical, optical and electronic properties that can be tuned by\naltering the number of carbon atoms, strain, encapsulation, and other external\nparameters. In this work, we probe the effects of quantum anharmonicity, strain\nand finite size on the structural and vibrational properties of these chains,\nusing high$-$level density functional theory (DFT) calculations. We find strong\nanharmonicity effects for infinite chains, leading to ground$-$state nuclear\nwavefunctions that are barely localized at each of the dimerized geometries,\ni.e. strong tunneling occurs between the two minima of the potential energy\nsurface. This effect is enhanced for compressive strains. In addition,\nvibrational C$-$band frequencies deviate substantially from experimental\nmeasurements in long chains encapsulated in carbon nanotubes. On the other\nhand, calculations for finite chains suggest that quantum anharmonicity effects\nare strongly suppressed in finite system, even in the extrapolation to the\ninfinite case. For finite systems, vibrational C$-$band frequencies agree well\nwith experimental values at zero pressure. However, these frequencies increase\nunder compressive strain, in contradiction with recent results. This\ncontradiction is not resolved by adding explicitly the encapsulating carbon\nnanotubes to our calculations. Our results indicate that LCCs embody an\nintriguing 1D system in which the behavior of very large finite systems do not\nreproduce or converge to the behavior of truly infinite ones.\n", " With the completion of human genome mapping, the focus of scientists seeking\nto explain the biological complexity of living systems is shifting from\nanalyzing the individual components (such as a particular gene or biochemical\nreaction) to understanding the set of interactions amongst the large number of\ncomponents that results in the different functions of the organism. To this\nend, the area of systems biology attempts to achieve a \"systems-level\"\ndescription of biology by focusing on the network of interactions instead of\nthe characteristics of its isolated parts. In this article, we briefly describe\nsome of the emerging themes of research in \"network\" biology, looking at\ndynamical processes occurring at the two different length scales of within the\ncell and between cells, viz., the intra-cellular signaling network and the\nnervous system. We show that focusing on the systems-level aspects of these\nproblems allows one to observe surprising and illuminating common themes\namongst them.\n", " This paper explores the genotype-phenotype relationship. It outlines\nconditions under which the dependence of a quantitative trait on the genome\nmight be predictable, based on measurement of a limited subset of genotypes. It\nuses the theory of real-valued Boolean functions in a systematic way to\ntranslate trait data into the Fourier domain. Important trait features, such as\nthe roughness of the trait landscape or the modularity of a trait have a simple\nFourier interpretation. Ruggedness at a gene location corresponds to high\nsensitivity to mutation, while a modular organization of gene activity reduces\nsuch sensitivity.\n Traits where rugged loci are rare will naturally compress gene data in the\nFourier domain, leading to a sparse representation of trait data, concentrated\nin identifiable, low-level coefficients. This Fourier representation of a trait\norganizes epistasis in a form which is isometric to the trait data. As Fourier\nmatrices are known to be maximally incoherent with the standard basis, this\npermits employing compressive sensing techniques to work from data sets that\nare relatively small -- sometimes even of polynomial size -- compared to the\nexponentially large sets of possible genomes.\n This theory provides a theoretical underpinning for systematic use of Boolean\nfunction machinery to dissect the dependency of a trait on the genome and\nenvironment.\n", " An optical mouse that is in common use with personal computers is employed to\nmeasure the motion of a pendulum. The pendulum can be monitored (i) realtime\nonly, or (ii) also with data storage for later detailed analysis using Excel.\nThe software developed for this purpose is a LabView executable algorithm. It\nallows the user to select among several modes that include filtering\noperations. The limiting resolution of this position sensor, which is in the\nneighborhood of 50 micrometers, is determined by the 'dpi' specification of the\nmouse.\n", " The COVID-19 pandemic has initiated an unprecedented worldwide effort to\ncharacterize its evolution through the mapping of mutations of the coronavirus\nSARS-CoV-2. The early identification of mutations that could confer adaptive\nadvantages to the virus, such as higher infectivity or immune evasion, is of\nparamount importance. However, the large number of currently available genomes\nprecludes the efficient use of phylogeny-based methods. Here we present\nCoVtRec, a fast and scalable Topological Data Analysis approach for the\nsurveillance of emerging adaptive mutations in large genomic datasets. Our\nmethod overcomes limitations of state-of-the-art phylogeny-based approaches by\nquantifying the potential adaptiveness of mutations merely by their topological\nfootprint in the genome alignment, without resorting to the reconstruction of a\nsingle optimal phylogenetic tree. Analyzing millions of SARS-CoV-2 genomes from\nGISAID, we find a correlation between topological signals and adaptation to the\nhuman host. By leveraging the stratification by time in sequence data, our\nmethod enables the high-resolution longitudinal analysis of topological signals\nof adaptation. We characterize the convergent evolution of the coronavirus\nthroughout the whole pandemic to date, report on emerging potentially adaptive\nmutations, and pinpoint mutations in Variants of Concern that are likely\nassociated with positive selection. Our approach can improve the surveillance\nof mutations of concern and guide experimental studies.\n", " We consider the model Poisson problem $-\\Delta u = f \\in \\Omega$, $u = g$ on\n$\\pa \\Omega$, where $ \\Omega $ is a bounded polyhedral domain in $\\RR^n$. The\nobjective of the paper is twofold. The first objective is to review the well\nposedness and the regularity of our model problem using appropriate weighted\nspaces for the data and the solution. We use these results to derive the domain\nof the Laplace operator with zero boundary conditions on a concave domain,\nwhich seems not to have been fully investigated before. We also mention some\nextensions of our results to interface problems for the Elasticity equation.\nThe second objective is to illustrate how anisotropic weighted regularity\nresults for the Laplace operator in 3D are used in designing efficient finite\nelement discretizations of elliptic boundary value problems, with the focus on\nthe efficient discretization of the Poisson problem on polyhedral domains in\n$\\RR^3$, following {\\em Numer. Funct. Anal. Optim.}, 28(7-8):775--824, 2007.\nThe anisotropic weighted regularity results described and used in the second\npart of the paper are a consequence of the well-posedness results in\n(isotropically) weighted Sobolev spaces described in the first part of the\npaper. The paper is based on the talk by the last named author at the Congress\nof Romanian Mathematicians, Brasov 2011, and is largely a survey paper.\n", " One-sided matching mechanisms are fundamental for assigning a set of\nindivisible objects to a set of self-interested agents when monetary transfers\nare not allowed. Two widely-studied randomized mechanisms in multiagent\nsettings are the Random Serial Dictatorship (RSD) and the Probabilistic Serial\nRule (PS). Both mechanisms require only that agents specify ordinal preferences\nand have a number of desirable economic and computational properties. However,\nthe induced outcomes of the mechanisms are often incomparable and thus there\nare challenges when it comes to deciding which mechanism to adopt in practice.\nIn this paper, we first consider the space of general ordinal preferences and\nprovide empirical results on the (in)comparability of RSD and PS. We analyze\ntheir respective economic properties under general and lexicographic\npreferences. We then instantiate utility functions with the goal of gaining\ninsights on the manipulability, efficiency, and envyfreeness of the mechanisms\nunder different risk-attitude models. Our results hold under various preference\ndistribution models, which further confirm the broad use of RSD in most\npractical applications.\n", " It is highly desirable to combine recent advances in the topological quantum\nphases with technologically relevant materials. Chromium dioxide (CrO2) is a\nhalf-metallic material, widely used in high-end data storage applications.\nUsing first principles calculations, we show via interfacial orbital design\nthat a novel class of half semi-metallic Dirac electronic phase emerges at the\ninterface CrO2 with TiO2 in both thin film and superlattice configurations,\nwith four spin-polarized Dirac points in momentum-space (k-space) band\nstructure. When the spin and orbital degrees of freedom are allowed to couple,\nthe CrO2/TiO2 superlattice becomes a Chern insulator without external fields or\nadditional doping. With topological gaps equivalent to 43 Kelvin and a Chern\nnumber 2, the ensuing quantization of Hall conductance to 2e^2/h will enable\npotential development of these highly industrialized oxides for applications in\ntopologically high fidelity data storage and energy-efficient electronic and\nspintronic devices.\n", " Scientific data has been growing in both size and complexity across the\nmodern physical, engineering, life and social sciences. Spatial structure, for\nexample, is a hallmark of many of the most important real-world complex\nsystems, but its analysis is fraught with statistical challenges. Topological\ndata analysis can provide a powerful computational window on complex systems.\nHere we present a framework to extend and interpret persistent homology\nsummaries to analyse spatial data across multiple scales. We introduce\nhyperTDA, a topological pipeline that unifies local (e.g. geodesic) and global\n(e.g. Euclidean) metrics without losing spatial information, even in the\npresence of noise. Homology generators offer an elegant and flexible\ndescription of spatial structures and can capture the information computed by\npersistent homology in an interpretable way. Here the information computed by\npersistent homology is transformed into a weighted hypergraph, where hyperedges\ncorrespond to homology generators. We consider different choices of generators\n(e.g. matroid or minimal) and find that centrality and community detection are\nrobust to either choice. We compare hyperTDA to existing geometric measures and\nvalidate its robustness to noise. We demonstrate the power of computing\nhigher-order topological structures on spatial curves arising frequently in\necology, biophysics, and biology, but also in high-dimensional financial\ndatasets. We find that hyperTDA can select between synthetic trajectories from\nthe landmark 2020 AnDi challenge and quantifies movements of different animal\nspecies, even when data is limited.\n", " The explosive growth of data results in more scarce spectrum resources. It is\nimportant to optimize the system performance under limited resources. In this\npaper, we investigate how to achieve weighted throughput (WTP) maximization for\ncell-free (CF) multiuser MIMO (MU-MIMO) multicarrier (MC) systems through\nresource allocation (RA), in the cases of finite blocklength (FBL) and infinite\nblocklength (INFBL) regimes. To ensure the quality of service (QoS) of each\nuser, particularly for the block error rate (BLER) and latency in the FBL\nregime, the WTP gets maximized under the constraints of total power consumption\nand required QoS metrics. Since the channels vary in different subcarriers\n(SCs) and inter-user interference strengths, the WTP can be maximized by\nscheduling the best users in each time-frequency (TF) resource and advanced\nbeamforming design, while the resources can be fully utilized. With this\nmotivation, we propose a joint user scheduling (US) and beamforming design\nalgorithm based on the successive convex approximation (SCA) and gene-aided\n(GA) algorithms, to address a mixed integer nonlinear programming (MINLP)\nproblem. Numerical results demonstrate that the proposed RA outperforms the\ncomparison schemes. And the CF system in our scenario is capable of achieving\nhigher spectral efficiency than the centralized antenna systems (CAS).\n", " The author researched this new idea - support of flight by any aerial\nvehicles at significant altitude solely by the magnetic field of the planet. It\nis shown that current technology allows humans to create a light propulsion (AB\nengine) which does not depend on air, water or ground terrain. Simultaniosly,\nthis revolutionary thruster is a device for the storage of electricity which is\nextracted and is replenished (during braking) from/into the storage with 100\npercent efficiency. The relative weight ratio of this engine is 0.01 - 0.1\n(from thrust). For some types of AB engine (toroidal form) the thrust easily\nmay be changed in any direction without turning of engine.\n The author computed many projects using different versions of offered AB\nengine: small device for levitation-flight of a human (including flight from\nEarth to Outer Space), fly VTOL car (track), big VTOL aircrat, suspended low\naltitude stationary satellite, powerful Space Shuttle-like booster for travel\nto the Moon and Mars without spending energy (spended energy is replenished in\nbraking when ship returns from other planet to its point of origin), using\nAB-devices in military, in sea-going ships (submarimes), in energy industry\n(for example. as small storage of electric energy) and so on. The vehicles\nequipped with AB propulsion can take flight for days and cover distances of\ntens thousands of kilometers at hypersonic or extra-atmosphere space speeds.\n The work contains tens of inventions and innovations which solves problems\nand breaks limitations which appear in solution of these very complex\nrevolutionary ideas.\n Key word: AB levitator, levitation, non-rocket outer space flight, electric\nenergy storage, AB propulsion, AB engine, Bolonkin.\n", " We study single spin asymmetries at one-loop accuracy in semi-inclusive DIS\nwith a transversely polarized hadron in the initial state. Two measurable spin\nobservables are predicted in the framework of QCD collinear factorization. One\nof the spin observables is the Sivers weighted asymmetry, another one is the\nCollins weighted asymmetry. The prediction takes a form of convolutions of\nperturbative coefficient functions and nonperturbative functions, which are\ntwist-2 transversity distributions, twist-3 parton distributions, twist-2- and\ntwist-3 parton fragmentation functions. These nonperturbative functions can be\nextracted from measurements of the spin observables and provide valuable\ninformation of the inner structure of hadrons. The measurements can be done in\ncurrent COMPASS- and JLab experiments and in future experiments of EIC. The\nperturbative coefficient functions are calculated at one-loop level. There are\ncollinear divergences in the calculation involving chirality-even- and\nchirality-odd twist operators. We find that all collinear divergences can be\ncorrectly subtracted so that the final results are finite.\n", " There are encouraging evidences for registration of excitations of quark\noscillator levels in compressed 2-nucleon systems, which were hidden in two\nindependently published experimental works for many years. Data obtained by EVA\ncollaboration can be considered as the third possible confirmation of this\nconjecture. Moreover, they drop hints about existence of coherent dibaryons\nwhich may be interpreted as usual or generalized coherent states of the\noscillator. These data may also be an explicit example of a physical process\ndescribed by Landau and Peierls in their reasoning concerning the energy-time\nuncertainty relation. Two models of the coherent dibaryon creation are\nformulated. Phase transitions of baryon matter in few-nucleon systems are\ndiscussed.\n", " We present a new method, Non-Stationary Forward Flux Sampling, that allows\nefficient simulation of rare events in both stationary and non-stationary\nstochastic systems. The method uses stochastic branching and pruning to achieve\nuniform sampling of trajectories in phase space and time, leading to accurate\nestimates for time-dependent switching propensities and time-dependent phase\nspace probability densities. The method is suitable for equilibrium or\nnon-equilibrium systems, in or out of stationary state, including non-Markovian\nor externally driven systems. We demonstrate the validity of the technique by\napplying it to a one-dimensional barrier crossing problem that can be solved\nexactly, and show its usefulness by applying it to the time-dependent switching\nof a genetic toggle switch.\n", " Probabilistic Boolean Networks have been proposed for estimating the\nbehaviour of dynamical systems as they combine rule-based modelling with\nuncertainty principles. Inferring PBNs directly from gene data is challenging\nhowever, especially when data is costly to collect and/or noisy, e.g., in the\ncase of gene expression profile data. In this paper, we present a reproducible\nmethod for inferring PBNs directly from real gene expression data measurements\ntaken when the system was at a steady state. The steady-state dynamics of PBNs\nis of special interest in the analysis of biological machinery. The proposed\napproach does not rely on reconstructing the state evolution of the network,\nwhich is computationally intractable for larger networks. We demonstrate the\nmethod on samples of real gene expression profiling data from a well-known\nstudy on metastatic melanoma. The pipeline is implemented using Python and we\nmake it publicly available.\n", " For polynomial ODE models, we introduce and discuss the concepts of exact and\napproximate conservation laws, which are the first integrals of the full and\ntruncated sets of ODEs. For fast-slow systems, truncated ODEs describe the fast\ndynamics. We define compatibility classes as subsets of the state space,\nobtained by equating the conservation laws to constants. A set of conservation\nlaws is complete when the corresponding compatibility classes contain a finite\nnumber of steady states. Complete sets of conservation laws can be used for\nmodel order reduction and for studying the multistationarity of the model. We\nprovide algorithmic methods for computing linear, monomial, and polynomial\nconservation laws of polynomial ODE models and for testing their completeness.\nThe resulting conservation laws and their completeness are either independent\nor dependent on the parameters. In the latter case, we provide parametric case\ndistinctions. In particular, we propose a new method to compute polynomial\nconservation laws by comprehensive Gr\\\"obner systems and syzygies.\n Keywords: First integrals, chemical reaction networks, polynomial\nconservation laws, syzygies, comprehensive Gr\\\"obner systems.\n", " The aim of this paper is to study the Wold-type decomposition in the class of\n$m$-isometries. One of our main results establishes an equivalent condition for\nan analytic $m$-isometry to admit the Wold-type decomposition for $m\\ge2$. In\nparticular, we introduce the $k$-kernel condition which we use to characterize\nanalytic $m$-isometric operators which are unitarily equivalent to unilateral\noperator valued weighted shifts for $m\\ge2$. As a result, we also show that\n$m$-isometric composition operators on directed graphs with one circuit\ncontaining only one element are not unitarily equivalent to unilateral weighted\nshifts. We also provide a characterization of $m$-isometric unilateral operator\nvalued weighted shifts with positive and commuting weights.\n", " The sensitivity of the displacement transducer pumped with a train of\nhigh-intensity laser pulses is estimated. Due to the multicomponent character\nof the pump a consideration of transformations of the signal and the noises\nbetween optical modes plays an important role in estimation of the potential\nsensitivity. An expression for the minimal detectable external classical force\nresembles those for the continuous wave pumping with substitution of the laser\npower by a time averaged power of pulsed laser. Possible scheme for back action\nnoise compensation for such transducers is considered. For full suppression of\nback action noise the field of local oscillator has to be pulsed with the same\ntime dependence as the pump field.\n", " In real applications, database systems should be able to manage and process\ndata with uncertainty. Any real dataset may have missing or rounded values,\nalso the values of data may change by time. So, it becomes important to handle\nthese uncertain data. An important problem in database technology is to cluster\nthese uncertain data.\n In this paper, we study the $k$-center problem for uncertain points in a\ngeneral metric space. First we present a greedy approximation algorithm that\nbuilds $k$ centers using a farthest-first traversal in $k$ iterations. This\nalgorithm improves the approximation factor of the unrestricted assigned\n$k$-center problem from $10$ to $6$. Next we restrict the centers to be\nselected from a finite set of points and we show that the optimal solution for\nthis restricted setting is a $2$-approximation factor solution for the optimal\nsolution of the assigned $k$-center problem. Using this idea we improve the\napproximation factor of the unrestricted assigned $k$-center problem to $4$ by\nincreasing the running time mildly.\n", " We present a novel approach for data set scaling based on scale-measures from\nformal concept analysis, i.e., continuous maps between closure systems, and\nderive a canonical representation. Moreover, we prove said scale-measures are\nlattice ordered with respect to the closure systems. This enables exploring the\nset of scale-measures through by the use of meet and join operations.\nFurthermore we show that the lattice of scale-measures is isomorphic to the\nlattice of sub-closure systems that arises from the original data. Finally, we\nprovide another representation of scale-measures using propositional logic in\nterms of data set features. Our theoretical findings are discussed by means of\nexamples.\n", " Chemical reactions occur in energy, environmental, biological, and many other\nnatural systems, and the inference of the reaction networks is essential to\nunderstand and design the chemical processes in engineering and life sciences.\nYet, revealing the reaction pathways for complex systems and processes is still\nchallenging due to the lack of knowledge of the involved species and reactions.\nHere, we present a neural network approach that autonomously discovers reaction\npathways from the time-resolved species concentration data. The proposed\nChemical Reaction Neural Network (CRNN), by design, satisfies the fundamental\nphysics laws, including the Law of Mass Action and the Arrhenius Law.\nConsequently, the CRNN is physically interpretable such that the reaction\npathways can be interpreted, and the kinetic parameters can be quantified\nsimultaneously from the weights of the neural network. The inference of the\nchemical pathways is accomplished by training the CRNN with species\nconcentration data via stochastic gradient descent. We demonstrate the\nsuccessful implementations and the robustness of the approach in elucidating\nthe chemical reaction pathways of several chemical engineering and biochemical\nsystems. The autonomous inference by the CRNN approach precludes the need for\nexpert knowledge in proposing candidate networks and addresses the curse of\ndimensionality in complex systems. The physical interpretability also makes the\nCRNN capable of not only fitting the data for a given system but also\ndeveloping knowledge of unknown pathways that could be generalized to similar\nchemical systems.\n", " We consider chemical reaction networks modeled by a discrete state and\ncontinuous in time Markov process for the vector copy number of the species and\nprovide a novel particle filter method for state and parameter estimation based\non exact observation of some of the species in continuous time. The conditional\nprobability distribution of the unobserved states is shown to satisfy a system\nof differential equations with jumps. We provide a method of simulating a\nprocess that is a proxy for the vector copy number of the unobserved species\nalong with a weight. The resulting weighted Monte Carlo simulation is then used\nto compute the conditional probability distribution of the unobserved species.\nWe also show how our algorithm can be adapted for a Bayesian estimation of\nparameters and for the estimation of a past state value based on observations\nup to a future time.\n", " Hanging about a hypothetical connections between the ground state vector for\nsome special spin systems and the alternating-sign matrices, we have found a\nnumerical evidence for the fact that the numbers of the states of the fully\npacked loop model with fixed link-patterns coincide with the components of the\nground state vector of the dense O$(1)$ loop model considered by Batchelor, de\nGier and Nienhuis. Our conjecture generalizes in a sense the conjecture of\nBosley and Fidkowski, refined by Cohn and Propp, and proved by Wieland.\n", " We are studying the mechanism of the cosmic model in the presence of GGPDE\nand matter in LRS Bianchi type-I space-time by the utilization of new\nholographic DE in Saez-Ballester theory. Here we discuss all the data for three\nscenarios, first is supernovae type Ia union data, second is SN Ia data in\ncombination with BAO and CMB observations and third is combination with OHD and\nJLA observations. From this, we get a model of our universe, where its transit\nstate from deceleration to acceleration phase. Here we have observed that the\nresults yielded by cosmological parameters like $\\rho$ (energy density), EoS\n(equation of state), squared speed of sound $(v_s^2)$,\n$(\\omega_D-\\omega_D^{'})$ and $(r-s)$ plane is compatible with the recent\nobservational data. The $(\\omega_D-\\omega_D^{'})$ trajectories in both thawing\nand freezing regions and the correspondence of the quintessence field with GGPD\ndark energy are discussed. Some physical aspects of the GGPDE models are also\nhighlighted.\n", " The distance function to a compact set plays a crucial role in the paradigm\nof topological data analysis. In particular, the sublevel sets of the distance\nfunction are used in the computation of persistent homology -- a backbone of\nthe topological data analysis pipeline. Despite its stability to perturbations\nin the Hausdorff distance, persistent homology is highly sensitive to outliers.\nIn this work, we develop a framework of statistical inference for persistent\nhomology in the presence of outliers. Drawing inspiration from recent\ndevelopments in robust statistics, we propose a $\\textit{median-of-means}$\nvariant of the distance function ($\\textsf{MoM Dist}$), and establish its\nstatistical properties. In particular, we show that, even in the presence of\noutliers, the sublevel filtrations and weighted filtrations induced by\n$\\textsf{MoM Dist}$ are both consistent estimators of the true underlying\npopulation counterpart, and their rates of convergence in the bottleneck metric\nare controlled by the fraction of outliers in the data. Finally, we demonstrate\nthe advantages of the proposed methodology through simulations and\napplications.\n", " Revealing the continuous dynamics on the networks is essential for\nunderstanding, predicting, and even controlling complex systems, but it is hard\nto learn and model the continuous network dynamics because of complex and\nunknown governing equations, high dimensions of complex systems, and\nunsatisfactory observations. Moreover, in real cases, observed time-series data\nare usually non-uniform and sparse, which also causes serious challenges. In\nthis paper, we propose an Autoregressive GNN-ODE GRU Model (AGOG) to learn and\ncapture the continuous network dynamics and realize predictions of node states\nat an arbitrary time in a data-driven manner. The GNN module is used to model\ncomplicated and nonlinear network dynamics. The hidden state of node states is\nspecified by the ODE system, and the augmented ODE system is utilized to map\nthe GNN into the continuous time domain. The hidden state is updated through\nGRUCell by observations. As prior knowledge, the true observations at the same\ntimestamp are combined with the hidden states for the next prediction. We use\nthe autoregressive model to make a one-step ahead prediction based on\nobservation history. The prediction is achieved by solving an initial-value\nproblem for ODE. To verify the performance of our model, we visualize the\nlearned dynamics and test them in three tasks: interpolation reconstruction,\nextrapolation prediction, and regular sequences prediction. The results\ndemonstrate that our model can capture the continuous dynamic process of\ncomplex systems accurately and make precise predictions of node states with\nminimal error. Our model can consistently outperform other baselines or achieve\ncomparable performance.\n", " We use a chiral random matrix model with 2N_f flavors to mock up the QCD\nDirac spectrum at finite chemical potential. We show that the 1/N approximation\nbreaks down in the quenched state with spontaneously broken chiral symmetry.\nThe breakdown condition is set by the divergence of a two-point function, that\nis shown to follow the general lore of macroscopic universality. In this state,\nthe fermionic fluctuations are not suppressed in the large N limit.\n", " The world is at the cusp of a new era where pivotal importance is being\nattached to the development of sustainable and high-performance energy storage\nsystems. Potassium-ion batteries are deemed not only as cheap battery\ncandidates, but also as the penultimate high-voltage energy storage systems\nwithin the monovalent-cation chemistries. However, their performance and\nsustainability are undermined by the lack of suitable electrolytes for\nhigh-voltage operation particularly due to the limited availability of cathode\nmaterials. Here, the potential of ionic liquids based on potassium\nbis(trifluoromethanesulfonyl)amide (KTFSA) as high-voltage electrolytes is\npresented by assessing their physicochemical properties, along with the\nelectrochemical properties upon coupling with new high-voltage layered cathode\nmaterials. These ionic liquids demonstrate a lower redox potential for\npotassium dissolution / deposition (with a wide voltage tolerance of around\n$6.0$ $\\rm V$), placing them as feasible and safe electrolytes for high-voltage\npotassium-ion battery configuration. This is proven by matching this\nelectrolyte with new high-voltage layered cathode compositions, demonstrating\nstable electrochemical performance. The present findings of electrochemically\nstable ionic liquids based on potassium bis(trifluoromethanesulfonyl)amide will\nbolster further advancement of high-performance cathode materials, whose\nperformance at high-voltage regimes were apparently restricted by the paucity\nof suitable and compatible electrolytes.\n", " We discuss basic features of emergent complexity in dynamical systems far\nfrom equilibrium by focusing on the network structure of their state space. We\nstart by measuring the distributions of avalanche and transient times in Random\nBoolean Networks (RBNs) and in the \\emph{Drosophila} polarity network by exact\nenumeration. A transient time is the duration of the transient from a starting\nstate to an attractor. An avalanche is a special transient which starts as\nsingle Boolean element perturbation of an attractor state. Significant\ndifferences at short times between the avalanche and the transient times for\nRBNs with small connectivity $K$ -- compared to the number of elements $N$ --\nindicate that attractors tend to cluster in configuration space. In addition,\none bit flip has a non-negligible chance to put an attractor state directly\nonto another attractor. This clustering is also present in the segment polarity\ngene network of \\emph{Drosophila melanogaster}, suggesting that this may be a\nrobust feature of biological regulatory networks. We also define and measure a\nbranching ratio for the state space networks and find evidence for a new time\nscale that diverges roughly linearly with $N$ for $2\\leq K \\ll N$. Analytic\narguments show that this time scale does not appear in the random map nor can\nthe random map exhibit clustering of attractors. We further show that for K=2\nthe branching ratio exhibits the largest variation with distance from the\nattractor compared to other values of $K$ and that the avalanche durations\nexhibit no characteristic scale within our statistical resolution. Hence, we\npropose that the branching ratio and the avalanche duration are new indicators\nfor scale-free behavior that may or may not be found simultaneously with other\nindicators of emergent complexity in extended, deterministic dynamical systems.\n", " Here we deconstruct, and then in a reasoned way reconstruct, the concept of\n\"entropy of a system,\" paying particular attention to where the randomness may\nbe coming from. We start with the core concept of entropy as a COUNT associated\nwith a DESCRIPTION; this count (traditionally expressed in logarithmic form for\na number of good reasons) is in essence the number of possibilities---specific\ninstances or \"scenarios,\" that MATCH that description. Very natural (and\nvirtually inescapable) generalizations of the idea of description are the\nprobability distribution and of its quantum mechanical counterpart, the density\noperator.\n We track the process of dynamically updating entropy as a system evolves.\nThree factors may cause entropy to change: (1) the system's INTERNAL DYNAMICS;\n(2) unsolicited EXTERNAL INFLUENCES on it; and (3) the approximations one has\nto make when one tries to predict the system's future state. The latter task is\nusually hampered by hard-to-quantify aspects of the original description,\nlimited data storage and processing resource, and possibly algorithmic\ninadequacy. Factors 2 and 3 introduce randomness into one's predictions and\naccordingly degrade them. When forecasting, as long as the entropy bookkeping\nis conducted in an HONEST fashion, this degradation will ALWAYS lead to an\nentropy increase.\n To clarify the above point we introduce the notion of HONEST ENTROPY, which\ncoalesces much of what is of course already done, often tacitly, in responsible\nentropy-bookkeping practice. This notion, we believe, will help to fill an\nexpressivity gap in scientific discourse. With its help we shall prove that ANY\ndynamical system---not just our physical universe---strictly obeys Clausius's\noriginal formulation of the second law of thermodynamics IF AND ONLY IF it is\ninvertible. Thus this law is a TAUTOLOGICAL PROPERTY of invertible systems!\n", " Metabolic networks perform some of the most fundamental functions in living\ncells, including energy transduction and building block biosynthesis. While\nthese are the best characterized networks in living systems, understanding\ntheir evolutionary history and complex wiring constitutes one of the most\nfascinating open questions in biology, intimately related to the enigma of\nlife's origin itself. Is the evolution of metabolism subject to general\nprinciples, beyond the unpredictable accumulation of multiple historical\naccidents? Here we search for such principles by applying to an artificial\nchemical universe some of the methodologies developed for the study of genome\nscale models of cellular metabolism. In particular, we use metabolic flux\nconstraint-based models to exhaustively search for artificial chemistry\npathways that can optimally perform an array of elementary metabolic functions.\nDespite the simplicity of the model employed, we find that the ensuing pathways\ndisplay a surprisingly rich set of properties, including the existence of\nautocatalytic cycles and hierarchical modules, the appearance of universally\npreferable metabolites and reactions, and a logarithmic trend of pathway length\nas a function of input/output molecule size. Some of these properties can be\nderived analytically, borrowing methods previously used in cryptography. In\naddition, by mapping biochemical networks onto a simplified carbon atom\nreaction backbone, we find that several of the properties predicted by the\nartificial chemistry model hold for real metabolic networks. These findings\nsuggest that optimality principles and arithmetic simplicity might lie beneath\nsome aspects of biochemical complexity.\n", " We introduce a mean counting function for Dirichlet series, which plays the\nsame role in the function theory of Hardy spaces of Dirichlet series as the\nNevanlinna counting function does in the classical theory. The existence of the\nmean counting function is related to Jessen and Tornehave's resolution of the\nLagrange mean motion problem. We use the mean counting function to describe all\ncompact composition operators with Dirichlet series symbols on the\nHardy--Hilbert space of Dirichlet series, thus resolving a problem which has\nbeen open since the bounded composition operators were described by Gordon and\nHedenmalm. The main result is that such a composition operator is compact if\nand only if the mean counting function of its symbol satisfies a decay\ncondition at the boundary of a half-plane.\n", " Multiple Reflection Expansion (MRE) formalism has been applied to hadron\nresonance gas (HRG) model to study the finite-size effect on thermodynamics of\nsmall systems of hadron gas at the chemical freeze-out temperature in\nhigh-multiplicity events of proton-proton (pp) colisions at the LHC. Comparison\nwith larger systems of heavy-ion (AA) collisions helps in undersanding the\nusefulness of the effect on small systems. Thermodynamic properties of these\nsystems at the chemical freeze-out, with and without system-size effect, are\ncontrasted with those for infinite hadronic phase of strongly interacting\nmatter at ideal thermodynamic limit, as provided by LQCD calculations. On\nintroduction of finite size effect, the small hadronic systems produced in\nhigh-multiplicity pp events, unlike those in AA collisions, remain away from\nideal thermodynamic limit. Knudsen number estimations validate the findings.\n", " Different space-filling curves (SFCs) are briefly reviewed in this paper, and\na new one is proposed. A century has passed between the inception of this kind\nof curves, since then they have been found useful in computer science,\nparticularly in data storage and indexing due to their clustering properties,\nbeing Hilbert curve the most well-known member of the family of fractals. The\nproposed Aztec curve, with similar characteristics to the Hilbert's curve, is\nintroduced in this paper, accompanied by a grammatical description for its\nconstruction. It yields the possibility of creating bi-dimensional clusters,\nnot available for Hilbert nor Peano curves. Additional to this, a case of\napplication on the scope of Compressed Sensing is implemented, in which the use\nof Hilbert curve is contrasted with Aztec curve, having a similar performance,\nand positioning the Aztec curve as viable and a new alternative for future\nexploitation on applications that make use of SFC's.\n", " Electron transport in realistic physical and chemical systems often involves\nthe non-trivial exchange of energy with a large environment, requiring the\ndefinition and treatment of open quantum systems. Because the time evolution of\nan open quantum system employs a non-unitary operator, the simulation of open\nquantum systems presents a challenge for universal quantum computers\nconstructed from only unitary operators or gates. Here we present a general\nalgorithm for implementing the action of any non-unitary operator on an\narbitrary state on a quantum device. We show that any quantum operator can be\nexactly decomposed as a linear combination of at most four unitary operators.\nWe demonstrate this method on a two-level system in both zero and finite\ntemperature amplitude damping channels. The results are in agreement with\nclassical calculations, showing promise in simulating non-unitary operations on\nintermediate-term and future quantum devices.\n" ]

[ " In this paper we discuss the $L^p$-$L^q$ boundedness of both spectral and\nFourier multipliers on general locally compact separable unimodular groups $G$\nfor the range $1<p\\leq q<\\infty$. We prove a Lizorkin type multiplier theorem\nfor $1<p\\leq q<\\infty$, and then refine it as a H\\\"ormander type multiplier\ntheorem for $1<p\\leq 2\\leq q<\\infty$. In the process, we establish versions of\nPaley and Hausdorff-Young-Paley inequalities on general locally compact\nseparable unimodular groups. As a consequence of the H\\\"ormander type\nmultiplier theorem we derive a spectral multiplier theorem on general locally\ncompact separable unimodular groups. We then apply it to obtain embedding\ntheorems as well as time-asymptotics for the $L^p$-$L^q$ norms of the heat\nkernels for general positive unbounded invariant operators on $G$. We\nillustrate the obtained results for sub-Laplacians on compact Lie groups and on\nthe Heisenberg group. We show that our results imply the known results for\n$L^p$-$L^q$ multipliers such as H\\\"ormander's Fourier multiplier theorem on\n$\\mathbb{R}^{n}$ or known results for Fourier multipliers on compact Lie\ngroups. The new approach developed in this paper relies on the analysis in the\ngroup von Neumann algebra for the derivation of the desired multiplier\ntheorems.\n", " Convolution semigroups of states on a quantum group form the natural\nnoncommutative analogue of convolution semigroups of probability measures on a\nlocally compact group. Here we initiate a theory of weakly continuous\nconvolution semigroups of functionals on a C*-bialgebra, the noncommutative\ncounterpart of locally compact semigroup. On locally compact quantum groups we\nobtain a bijective correspondence between such convolution semigroups and a\nclass of C_0-semigroups of maps which we characterise. On C*-bialgebras of\ndiscrete type we show that all weakly continuous convolution semigroups of\nstates are automatically norm-continuous. As an application we deduce a known\ncharacterisation of continuous conditionally positive-definite Hermitian\nfunctions on a compact group.\n", " We prove an almost everywhere convergence result for Bochner-Riesz means of\n$L^p$ functions on Heisenberg-type groups, yielding the existence of a $p>2$\nfor which convergence holds for means of arbitrarily small order. The proof\nhinges on a reduction of weighted $L^2$ estimates for the maximal Bochner-Riesz\noperator to corresponding estimates for the non-maximal operator, and a `dual\nSobolev trace lemma', whose proof is based on refined estimates for Jacobi\npolynomials.\n", " The boundary behaviour of convolutions with Poisson kernel and with square\nroot from Poisson kernel is essentially differs. The first ones have only\nnontangential limit. For the last ones the convergence is over domains\nadmittings a logarithmic order of the contact with the boundary . This result\nwas generalized by authors on the spaces of homogeneous type.\n Here we prove the boundedness and some weighted estimates for the\ncorresponding maximal operator. Earlier it was known only weak type inequality.\n", " Using the ideas of E.I. Gordon we present and farther advance an approach,\nbased on nonstandard analysis, to simultaneous approximations of locally\ncompact abelian groups and their duals by (hyper)finite abelian groups, as well\nas to approximations of various types of Fourier transforms on them by the\ndiscrete Fourier transform. Combining some methods of nonstandard analysis and\nadditive combinatorics we prove the three Gordon's Conjectures which were open\nsince 1991 and are crucial both in the formulations and proofs of the LCA\ngroups and Fourier transform approximation theorems.\n", " We give conditions for boundedness of Hausdorff operators on real Hardy\nspaces $H^1$ over homogeneous spaces of locally compact groups with local\ndoubling property. The special case of the hyperbolic plane is considered.\n", " We give a proof of the existence of generalized definable locally compact\nmodels for arbitrary approximate subgroups via an application of topological\ndynamics in model theory. Our construction is simpler and shorter than the\noriginal one obtained by Hrushovski in ``Beyond the Lascar group'', and it uses\nonly basic model theory (mostly spaces of types and realizations of types). The\nmain tools are Ellis groups from topological dynamics considered for suitable\nspaces of types. However, we need to redevelop some basic theory of topological\ndynamics for suitable ``locally compact flows'' in place of (compact) flows. We\nalso prove that the generalized definable locally compact model which we\nconstructed is universal in an appropriate category. We note that the main\nresult yields structural information on definable generic subsets of definable\ngroups, with a more precise structural result for generics in the universal\ncover of $\\textrm{SL}_2(\\mathbb{R})$.\n", " We develop a semigroup approach to representation theory for pro-Lie groups\nsatisfying suitable amenability conditions. As an application of our approach,\nwe establish a one-to-one correspondence between equivalence classes of unitary\nirreducible representations and coadjoint orbits for a class of pro-Lie groups\nincluding all connected locally compact nilpotent groups and arbitrary infinite\ndirect products of nilpotent Lie groups. The usual $C^*$-algebraic approach to\ngroup representation theory positivey breaks down for infinite direct products\nof non-compact locally compact groups, hence the description of their unitary\nduals in terms of coadjoint orbits is particularly important whenever it is\navailable, being the only description known so far.\n", " We give sufficient conditions on the Lebesgue exponents for compositions of\nodd numbers of pseudo-differential operators with symbols in modulation spaces.\nAs a byproduct, we obtain sufficient conditions for twisted convolutions of odd\nnumbers of factors to be bounded on Wiener amalgam spaces.\n", " For a left action $S\\overset{\\lambda}{\\curvearrowright}X$ of a cancellative\nright amenable monoid $S$ on a discrete Abelian group $X$, we construct its Ore\nlocalization $G\\overset{\\lambda^*}{\\curvearrowright}X^*$, where $G$ is the\ngroup of left fractions of $S$; analogously, for a right action\n$K\\overset{\\rho}\\curvearrowleft S$ on a compact space $K$, we construct its Ore\ncolocalization $K^*\\overset{\\rho^*}{\\curvearrowleft} G$. Both constructions\npreserve entropy, i.e., for the algebraic entropy $h_{\\mathrm{alg}}$ and for\nthe topological entropy $h_{\\mathrm{top}}$ one has\n$h_{\\mathrm{alg}}(\\lambda)=h_{\\mathrm{alg}}(\\lambda^*)$ and\n$h_{\\mathrm{top}}(\\rho)=h_{\\mathrm{top}}(\\rho^*)$, respectively.\n Exploiting these constructions and the theory of quasi-tilings, we extend the\nAddition Theorem for $h_{\\mathrm{top}}$, known for right actions of countable\namenable groups on compact metrizable groups, to right actions\n$K\\overset{\\rho}{\\curvearrowleft} S$ of cancellative right amenable monoids $S$\n(with no restrictions on the cardinality) on arbitrary compact groups $K$.\n When the compact group $K$ is Abelian, we prove that $h_{\\mathrm{top}}(\\rho)$\ncoincides with $h_{\\mathrm{alg}}(\\hat{\\rho})$, where\n$S\\overset{\\hat{\\rho}}\\curvearrowright X$ is the dual left action on the\ndiscrete Pontryagin dual $X=\\hat{K}$, that is, a so-called Bridge Theorem. From\nthe Addition Theorem for $h_{\\mathrm{top}}$ and the Bridge Theorem, we obtain\nan Addition Theorem for $h_{\\mathrm{alg}}$ for left actions\n$S\\overset{\\lambda}\\curvearrowright X$ on discrete Abelian groups, so far known\nonly under the hypotheses that either $X$ is torsion or $S$ is locally\nmonotileable.\n The proofs substantially use the unified approach towards entropy based on\nthe entropy of actions of cancellative right amenable monoids on appropriately\ndefined normed monoids.\n", " This paper provides a self-contained exposition of coorbit spaces associated\nto integrable group representations and quasi-Banach function spaces, and at\nthe same time extends and simplifies previous work. The main results provide an\nextension of the theory in [Studia Math., 180(3):237-253, 2007] from groups\nadmitting a compact, conjugation-invariant unit neighborhood to arbitrary\n(possibly nonunimodular) locally compact groups. In addition, the present paper\nestablishes the existence of molecular dual frames and Riesz sequences as in\n[J. Funct. Anal., 280(10):56, 2021] for the full scale of quasi-Banach function\nspaces. The theory is developed for possibly projective and reducible unitary\nrepresentations in order to be easily applicable to well-studied function\nspaces not satisfying the classical assumptions of coorbit theory. Compared to\nthe existing literature on quasi-Banach coorbit spaces, all our results apply\nunder significantly weaker integrability conditions on the analyzing vectors,\nwhich allows for obtaining sharp results in concrete settings\n", " We study topological groups $G$ for which the universal minimal $G$-system\n$M(G)$, or the universal irreducible affine $G$-system $IA(G)$ are tame. We\ncall such groups intrinsically tame and convexly intrinsically tame. These\nnotions are generalized versions of extreme amenability and amenability,\nrespectively. When $M(G)$, as a $G$-system, admits a circular order we say that\n$G$ is intrinsically circularly ordered. This implies that $G$ is intrinsically\ntame. We show that for every circularly ultrahomogeneous action $G\n\\curvearrowright X$ on a circularly ordered set $X$ the topological group $G$,\nin its pointwise convergence topology, is intrinsically circularly ordered.\nThis result is a \"circular\" analog of Pestov's result about the extremal\namenability of ultrahomogeneous actions on linearly ordered sets by linear\norder preserving transformations. In the case where $X$ is countable, the\ncorresponding Polish group of circular automorphisms $G$ admits a concrete\ndescription. Using the Kechris-Pestov-Todorcevic construction we show that\n$M(G)$ is a circularly ordered compact space obtained by splitting the rational\npoints on the circle. We show also that $G$ is Roelcke precompact, satisfies\nKazhdan's property $T$ (using results of Evans-Tsankov) and has the automatic\ncontinuity property (using results of Rosendal-Solecki).\n", " Contents:\n 1. Combinatorial and model-theoretical principles related to regularity of\nultrafilters and compactness of topological spaces, I; 2. Frechet-Urysohn fans\nin free topological groups; 3. Packing index of subsets in Polish groups; 4.\nSymmetric monochromatic subsets in colorings of the Lobachevsky plane; 5.\nStructural Ramsey theory of metric spaces and topological dynamics of isometry\ngroups; 6. Distinguishing Number of Countable Homogeneous Relational\nStructures; 7. Indestructible colourings and rainbow Ramsey theorems; 8.\nProducts of Borel subgroups; 9. Selection theorems and treeability; 10.\nCombinatorial and model-theoretical principles related to regularity of\nultrafilters and compactness of topological spaces, IV; 11. A property of Cp[0,\n1]; 12. A Dedekind Finite Borel Set; 13. Aronszajn Compacta; 14. A strong\nantidiamond principle compatible with CH; 15. On the strength of Hausdorff's\ngap condition; 16. Nonhomogeneous analytic families of trees; 17. Reasonable\nnon-Radon-Nikodym ideals; 18. Continuity and related forcing; 19. An exact\nRamsey principle for block sequences; 20. Baire reflection; 21. Tukey classes\nof ultrafilters on; 22. Countably determined compact abelian groups; 23. A\ntopological reflection principle equivalent to Shelah's Strong Hypothesis; 24.\nSuperfilters, Ramsey theory, and van der Waerden's Theorem.\n", " We prove that the family of normal currents in the sense of Rumin in a Carnot\ngroup is compact in the flat topology. This result is obtained through a dual\ncompactness argument for Rumin forms, using the pseudo-differential calculus in\ngroups developed by Folland, Christ-Geller-G lowacki-Polin and Rumin. As an\napplication, imitating de Pauw-Moonens-Pfeffer, we describe the space of\ncharges on a Carnot group.\n", " We prove that if open subgroups of the groups of invertible elements in two\nFourier-Stieltjes algebras are isometric as metric spaces, then the underlying\nlocally compact groups are topologically isomorphic. We describe the structure\nof isometric real algebra isomorphisms between Fourier-Stieltjes algebras and\napply it to prove the above result.\n", " Online principal component analysis (PCA) has been an efficient tool in\npractice to reduce dimension. However, convergence properties of the\ncorresponding ODE are still unknown, including global convergence, stable\nmanifolds, and convergence rate. In this paper, we focus on the stochastic\ngradient ascent (SGA) method proposed by Oja. By regarding the corresponding\nODE as a Landau-Lifshitz-Gilbert (LLG) equation on the Stiefel manifold, we\nproved global convergence of the ODE. Moreover, we developed a new technique to\ndetermine stable manifolds. This technique analyzes the rank of the initial\ndatum. Using this technique, we derived the explicit expression of the stable\nmanifolds. As a consequence, exponential convergence to stable equilibrium\npoints was also proved. The success of this new technique should be attributed\nto the semi-decoupling property of the SGA method: iteration of previous\ncomponents does not depend on that of later ones. As far as we know, our result\nis the first complete one on the convergence of an online PCA flow, providing\nglobal convergence, explicit characterization of stable manifolds, and closed\nformula of exponential convergence depending on the spectrum gap.\n", " In this article, we discuss subspace duals of a frame of translates by an\naction of a closed abelian subgroup $\\Gamma$ of a locally compact group\n$\\mathscr G.$ These subspace duals are not required to lie in the space\ngenerated by the frame. We characterise translation-generated subspace duals of\na frame/Riesz basis involving the Zak transform for the pair $(\\mathscr G,\n\\Gamma) .$ We continue our discussion on the orthogonality of two\ntranslation-generated Bessel pairs using the Zak transform, which allows us to\nexplore the dual of super-frames. As an example, we extend our findings to\nsplines, Gabor systems, $p$-adic fields $\\mathbb Q p,$ locally compact abelian\ngroups using the fiberization map.\n", " Let L be a positive line bundle over a compact complex projective manifold X\nand K be a compact subset of X which is regular in a sense of pluripotential\ntheory. A Fekete configuration of order k is a finite subset of K maximizing a\nVandermonde type determinant associated with the power L^k of L. Berman,\nBoucksom and Witt Nystr\\\"om proved that the empirical measure associated with a\nFekete configuration converges to the equilibrium measure of K as k tends to\ninfinity. Dinh, Ma and Nguyen obtained an estimate for the rate of convergence.\nUsing techniques from Cauchy-Riemann geometry, we show that the last result\nholds when K is a real nondegenerate C^5-piecewise submanifold of X such that\nits tangent space at any regular point is not contained in a complex hyperplane\nof the tangent space of X at that point. In particular, the estimate holds for\nFekete points on some compact sets in R^n or the unit sphere in R^{n+1}.\n", " The paper deals with $L_p$-boundedness of the Hartley-Fourier convolutions\noperator and their applied aspects. We establish various new Young-type\ninequalities and obtain the structure of a normed ring in Banach space when\nequipping it with such convolutional multiplication. Weighted $L_p$-norm\ninequalities of these convolutions are also considered. As applications, we\ninvestigate the solvability and the bounded $L_1$-solution of a class of\nFredholm-type integral equations and linear Barbashin's equations with the help\nof factorization identities of such convolutions. Several examples are provided\nto illustrate the obtained results to ensure their validity and applicability.\n", " This paper shows that a formula for the pushforward measures on the fiber\nbundle $ G \\to G / H $ of homogeneous space of locally compact groups $ G $ and\n$ H $ can be written down by using the modular functions. As a result, we\nobtain the equality \\begin{align} \\frac{ \\Delta_G ( h ) }{ \\Delta_{ G / N } ( h\nN ) } = \\frac{ \\Delta_H ( h ) }{ \\Delta_{ H / N } ( h N ) } \\end{align} on the\nmodular functions for any closed subgroups $ N $ and $ H $ of a locally compact\ngroup $ G $ with $ N \\lhd G $ and $ N \\subset H \\subset G $ and any $ h \\in H\n$.\n", " We give sharp asymptotic estimates at infinity of all radial partial\nderivatives of the heat kernel on H-type groups. As an application, we give a\nnew proof of the discreteness of the spectrum of some natural sub-Riemannian\nOrnstein-Uhlenbeck operators on these groups.\n", " Several recent papers investigated unbounded convergences in Banach lattices.\nCombine all unbounded convergences, including unbounded order (norm, absolute\nweak, absolute weak*) convergence, we characterize L-weakly compact sets,\nL-weakly compact operators and M-weakly compact operators on Banach lattices.\nSome related results are obtained as well.\n", " We consider the question whether, given a countable system of lattices\n$(\\Gamma_j)_{j \\in J}$ in a locally compact abelian group $G$, there exists a\nsequence of functions $(g_j)_{j \\in J}$ such that the resulting generalized\nshift-invariant system $(g_j(\\cdot - \\gamma))_{j \\in J, \\gamma \\in \\Gamma_j}$\nis a tight frame of $L^2(G)$. This paper develops a new approach to the study\nof almost periodic functions for generalized shift-invariant systems based on\nan \\emph{unconditionally convergence property}, replacing previously used local\nintegrability conditions. From this theory, we derive characterizing relations\nfor tight and dual frame generators, we introduce the \\emph{system bandwidth}\nas a measure of the total bandwidth a generalized shift-invariant system can\ncarry, and we show that the so-called Calder\\'on sum is uniformly bounded from\nbelow for generalized shift-invariant frames. We exhibit a condition on the\nlattice system for which the unconditionally convergence property is guaranteed\nto hold. Without the unconditionally convergence property, we show, counter\nintuitively, that even orthonormal bases can have arbitrary small system\nbandwidth. Our results show that the question of existence of frame generators\nfor a general lattice system can be rather subtle, depending on analytical\nproperties, such as the system bandwidth, as well as on algebraic properties of\nthe lattice system.\n", " Since the publication of the important work of Rauch and Taylor (Potential\nand scattering theory on wildly perturbed domains, JFA, 1975) a lot has been\ndone to analyse wild perturbations of the Laplace operator. Here we present\nresults concerning the norm convergence of the resolvent. We consider a (not\nnecessarily compact) manifold with many small balls removed, the number of\nballs can increase as the radius is shrinking, the number of balls can also be\ninfinite. If the distance of the balls shrinks less fast than the radius, then\nwe show that the Neumann Laplacian converges to the unperturbed Laplacian,\ni.e., the obstacles vanish. In the Dirichlet case, we have two cases: if the\nballs are too sparse, the limit operator is again the unperturbed one, while if\nthe balls concentrate at a certain region (they become \"solid\" in a region),\nthe limit operator is the Dirichlet Laplacian on the complement outside the\nsolid region. Our work is based on a norm convergence result for operators\nacting in varying Hilbert spaces developed by the second author.\n", " Many infinite-dimensional Lie groups of interest can be expressed as a union\nof an ascending sequence of (finite- or infinite-dimensional) Lie groups. In\nthis survey article, we compile general results concerning such ascending\nunions, describe the main classes of examples, and explain what the general\ntheory tells us about these.\n In particular, we discuss:\n (1) Direct limit properties of ascending unions of Lie groups in the relevant\ncategories;\n (2) Regularity in Milnor's sense;\n (3) Homotopy groups of direct limit groups and of Lie groups containing a\ndense union of Lie groups;\n (4) Subgroups of direct limit groups;\n (5) Constructions of Lie group structures on ascending unions of Lie groups.\n", " The full lattice convergence on a locally solid Riesz space is an abstraction\nof the topological, order, and relatively uniform convergences. We investigate\nfour modifications of a full convergence $\\mathbb{c}$ on a Riesz space. The\nfirst one produces a sequential convergence $\\mathbb{sc}$. The second makes an\nabsolute $\\mathbb{c}$-convergence and generalizes the absolute weak\nconvergence. The third modification makes an unbounded $\\mathbb{c}$-convergence\nand generalizes various unbounded convergences recently studied in the\nliterature. The last one is applicable whenever $\\mathbb{c}$ is a full\nconvergence on a commutative $l$-algebra and produces the multiplicative\nmodification $\\mathbb{mc}$ of $\\mathbb{c}$. We study general properties of full\nlattice convergence with emphasis on universally complete Riesz spaces and on\nArchimedean $f$-algebras. The technique and results in this paper unify and\nextend those which were developed and obtained in recent literature on\nunbounded convergences.\n", " In this paper we study the characters of sequences of representations of any\nof the three families of classical Weyl groups W_n: the symmetric groups, the\nsigned permutation groups (hyperoctahedral groups), or the even-signed\npermutation groups. Our results extend work of Church, Ellenberg, Farb, and\nNagpal on the symmetric groups. We use the concept of an FI_W-module, an\nalgebraic object that encodes the data of a sequence of W_n-representations\nwith maps between them, defined in the author's recent work ArXiv:1309.3817.\n We show that if a sequence {V_n} of W_n-representations has the structure of\na finitely generated FI_W-module, then there are substantial constraints on the\ngrowth of the sequence and the structure of the characters: for n large, the\ndimension of V_n is equal to a polynomial in n, and the characters of V_n are\ngiven by a character polynomial in signed-cycle-counting class functions,\nindependent of n. We determine bounds the degrees of these polynomials.\n We continue to develop the theory of FI_W-modules, and we apply this theory\nto obtain new results about a number of sequences associated to the classical\nWeyl groups: the cohomology of complements of classical Coxeter hyperplane\narrangements, and the cohomology of the pure string motion groups (the groups\nof symmetric automorphisms of the free group).\n", " In this paper, we obtain Hardy, Hardy-Rellich and refined Hardy inequalities\non general stratified groups and weighted Hardy inequalities on general\nhomogeneous groups using the factorization method of differential operators,\ninspired by the recent work of Gesztesy and Littlejohn \\cite{GL17}. We note\nthat some of the obtained inequalities are new also in the usual Euclidean\nsetting. We also obtain analogues of Gesztesy and Littlejohn's 2-parameter\nversion of the Rellich inequality on stratified groups and on the Heisenberg\ngroup, and a new two-parameter estimate on $\\mathbb{R}^{n}$ which can be\nregarded as a counterpart to the Gesztesy and Littlejohn's estimate.\n", " We prove a multiplier theorem for certain Laplacians with drift on\nDamek-Ricci spaces, which are a class of Lie groups of exponential growth. Our\ntheorem generalizes previous results obtained by W. Hebisch, G. Mauceri and S.\nMeda on Lie groups of polynomial growth.\n", " Two Schubert problems on possibly different Grassmannians may be composed to\nobtain a Schubert problem on a larger Grassmannian whose number of solutions is\nthe product of the numbers of the original problems. This generalizes a\nconstruction discovered while classifying Schubert problems on the Grassmannian\nof 4-planes in C^9 with imprimitive Galois groups. We give an algebraic proof\nof the product formula. In a number of cases, we show that the Galois group of\nthe composed Schubert problem is a subgroup of a wreath product of the Galois\ngroups of the original problems, and is therefore imprimitive. We also present\nevidence for a conjecture that all composed Schubert problems have imprimitive\nGalois groups.\n", " Topological measures and deficient topological measures generalize Borel\nmeasures and correspond to certain non-linear functionals. We study integration\nwith respect to deficient topological measures on locally compact spaces. Such\nan integration over sets yields a new deficient topological measure if we\nintegrate a nonnegative vanishing at infinity function; and it produces a\nsigned deficient topological measure if we use a continuous function on a\ncompact space. We present many properties of these resulting deficient\ntopological measures and of signed deficient topological measures. In\nparticular, they are absolutely continuous with respect to the original\ndeficient topological measure and Lipschitz continuous. Deficient topological\nmeasures obtained by integration over sets can also be obtained from non-linear\nfunctionals. We show that for a deficient topological measure $ \\mu$ that\nassumes finitely many values, there is a function $ f $ such that $\\int_X f \\,\nd \\mu = 0$, but $\\int_X (-f )\\, d \\mu \\neq 0$. We present different criteria\nfor $\\int_X f \\, d \\mu = 0$. We also prove some convergence results, including\na Monotone convergence theorem.\n", " We provide a general product formula for the solution of nonautonomous\nabstract delay equations. After having shown the convergence we obtain\nestimates on the order of convergence for differentiable history functions.\nFinally, the theoretical results are demonstrated on some typical numerical\nexamples.\n", " The aim of this paper is to investigate the quality of approximation of\nalmost time and band limited functions by its expansion in the Hermite and\nscaled Hermite basis. As a corollary, this allows us to obtain the rate of\nconvergence of the Hermite expansion of function in the $L^2$-Sobolev space\nwith fixed compact support.\n", " We focus on quantum systems represented by a Hilbert space $L^2(A)$, where\n$A$ is a locally compact Abelian group that contains a compact open subgroup.\nWe examine two interconnected issues related to Weyl-Heisenberg operators.\nFirst, we provide a complete and elegant solution to the problem of describing\nthe stabilizer states in terms of their wave functions, an issue that arises in\nquantum information theory. Subsequently, we demonstrate that the stabilizer\nstates are precisely the minimizers of the Wehrl entropy functional, thereby\nresolving the analog of the Wehrl conjecture for any such group. Additionally,\nwe construct a moduli space for the set of stabilizer states, that is, a\nparameterization of this set, that endows it with a natural algebraic\nstructure, and we derive a formula for the number of stabilizer states when $A$\nis finite. Notably, these results are novel even for finite Abelian groups.\n", " We express the discrete Ricci curvature of a graph as the minimal eigenvalue\nof a family of matrices, one for each vertex of a graph whose entries depend on\nthe local adjaciency structure of the graph. Using this method we compute or\nbound the Ricci curvature of Cayley graphs of finite Coxeter groups and affine\nWeyl groups. As an application we obtain an isoperimetric inequality that holds\nfor all Cayley graphs of finite Coxeter groups.\n", " In this paper, we present a topology optimization (TO) framework to enable\nautomated design of mechanical components while ensuring the result can be\nmanufactured using multi-axis machining. Although TO improves the part's\nperformance, the as-designed model is often geometrically too complex to be\nmachined and the as-manufactured model can significantly vary due to machining\nconstraints that are not accounted for during TO. In other words, many of the\noptimized design features cannot be accessed by a machine tool without\ncolliding with the part (or fixtures). The subsequent post-processing to make\nthe part machinable with the given setup requires trial-and-error without\nguarantees on preserving the optimized performance. Our proposed approach is\nbased on the well-established accessibility analysis formulation using\nconvolutions in configuration space that is extensively used in spatial\nplanning and robotics. We define an 'inaccessibility measure field' (IMF) over\nthe design domain to identify non-manufacturable features and quantify their\ncontribution to non-manufacturability. The IMF is used to penalize the\nsensitivity field of performance objectives and constraints to prevent\nformation of inaccessible regions. Unlike existing discrete formulations, our\nIMF provides a continuous spatial field that is desirable for TO convergence.\nOur approach applies to arbitrary geometric complexity of the part, tools, and\nfixtures, and is highly parallelizable on multi-core architecture. We\ndemonstrate the effectiveness of our framework on benchmark and realistic\nexamples in 2D and 3D. We also show that it is possible to directly construct\nmanufacturing plans for the optimized designs based on the accessibility\ninformation.\n", " We prove a Freiman-type theorem for locally compact abelian groups. If A is a\nsubset of a locally compact abelian group with Haar measure m and m(nA) <\nn^dm(A) for all n>d log d then we describe A in a way which is tight up to\nlogarithmic factors in d.\n", " Let (k(n)) n=1,2,... be a strictly increasing sequence of positive integers .\nWe consider a specific sequence of differential operators Tk(n),{\\lambda} ,\nn=1,2,... on the space of entire functions , that depend on the sequence (k(n))\nn=1,2,... and the non-zero complex number {\\lambda} . We establish the\nexistence of an entire function f , such that for every positive number\n{\\lambda} the set {Tk(n),{\\lambda} ,n=1,2,... } is dense in the space of entire\nfunctions endowed with the topology of uniform convergence on compact subsets\nof the complex plane . This provides the best possible strenghthened version of\na corresponding result due to Costakis and Sambarino [9] . From this and using\na non-trivial result of Weyl which concerns the uniform distribution modulo 1\nof certain sequences and Cavalieri principle we can extend our result for a\nsubset of the set of complex numbers with full 2-dimentional Lebesque measure .\n", " In this paper we consider the so-called crystallographic Radon transform (or\ncrystallographic $X$-ray transform) and totally geodesic Radon transform on the\ngroup of rotations SO(3). As we show both of these transforms naturally appear\nin texture analysis, i.e. the analysis of preferred crystallographic\norientation. Although we discuss only applications to texture analysis both\ntransforms have other applications as well.\n In section 2 we start with motivations and applications. In sections 3 and 4\nwe develop a general framework on compact Lie groups. In section 5 we give a\ndetailed analysis of the totally geodesic Radon transform on SO(3). In section\n\\ref{relations} we compare crystallographic Radon transform on SO(3) and Funk\ntransform on $S^{3}$. In section \\ref{1} we show non-invertibility of the\ncrystallographic transform. In section 8 we describe an exact reconstruction\nformula for bandlimited functions, which uses only a finite number of samples\nof their Radon transform. Some auxiliary results for this section are collected\nin Appendix.\n", " We study the following question: given a locally compact group when does its\nFourier algebra coincide with the subalgebra of the Fourier-Stieltjes algebra\nconsisting of functions vanishing at infinity? We provide sufficient conditions\nfor this to be the case.\n As an application, we show that when P is the minimal parabolic subgroup in\none of the classical simple Lie groups of real rank one or the exceptional such\ngroup, then the Fourier algebra of P coincides with the subalgebra of the\nFourier-Stieltjes algebra of P consisting of functions vanishing at infinity.\nIn particular, the regular representation of P decomposes as a direct sum of\nirreducible representations although P is not compact.\n", " We create a new, functional calculus, approach to approximation of\nC_0-semigroups on Banach spaces. As an application of this approach, we obtain\noptimal convergence rates in classical approximation formulas for\nC_0-semigroups. In fact, our methods allow one to derive a number of similar\nformulas and equip them with sharp convergence rates. As a byproduct, we prove\na new interpolation principle leading to efficient norm estimates in the Banach\nalgebra of Laplace transforms of bounded measures on the semi-axis.\n", " In our present investigation, we are concerned with the Kantorovich variant\nof Lupa\\c{s}-Stancu operators based on Polya distribution with Pochhammer\n$k$-symbol. We briefly give some basic properties of the generalized operators\nand by making use of these results, we investigate convergence properties of\nthe studied operators. Furthermore, the rate of convergence of these operators\nis obtained and Voronovskaja type theorem for the pointwise approximation is\nestablished. Then we construct bivariate generalization of the operators and we\ndiscuss some convergence properties. Finally, taking into account some\nillustrative graphics, we conclude our study with the comparison of the rate of\nconvergence between our operators and other operators which are mentioned in\nthe paper.\n", " We study $\\epsilon$-representations of discrete groups by unitary operators\non a Hilbert space. We define the notion of Ulam stability of a group which\nloosely means that finite-dimensional $\\epsilon$-represendations are uniformly\nclose to unitary representations. One of our main results is that certain\nlattices in connected semi-simple Lie groups of higher rank are Ulam stable.\nFor infinite-dimensional $\\epsilon$-representations, the similarly defined\nnotion of strong Ulam stability is defined and it is shown that groups with\nfree subgroups are not strongly Ulam stable. We also study deformation rigidity\nof unitary representations and show that groups containing a free subgroup are\nnot deformation rigid.\n", " We study uniform and coarse embeddings between Banach spaces and topological\ngroups. A particular focus is put on equivariant embeddings, i.e., continuous\ncocycles associated to continuous affine isometric actions of topological\ngroups on separable Banach spaces with varying geometry.\n", " The persistent homology with coefficients in a field F coincides with the\nsame for cohomology because of duality. We propose an implementation of a\nrecently introduced algorithm for persistent cohomology that attaches\nannotation vectors with the simplices. We separate the representation of the\nsimplicial complex from the representation of the cohomology groups, and\nintroduce a new data structure for maintaining the annotation matrix, which is\nmore compact and reduces substancially the amount of matrix operations. In\naddition, we propose heuristics to simplify further the representation of the\ncohomology groups and improve both time and space complexities. The paper\nprovides a theoretical analysis, as well as a detailed experimental study of\nour implementation and comparison with state-of-the-art software for persistent\nhomology and cohomology.\n", " Garonzi and Lucchini~\\cite{GL} explored finite groups $G$ possessing a normal\n$2$-covering, where no proper quotient of $G$ exhibits such a covering. Their\ninvestigation offered a comprehensive overview of these groups, delineating\nthat such groups fall into distinct categories: almost simple, affine, product\naction, or diagonal.\n In this paper, we focus on the family falling under the diagonal type.\nSpecifically, we present a thorough classification of finite diagonal groups\npossessing a normal $2$-covering, with the attribute that no proper quotient of\n$G$ has such a covering.\n", " The cubature on Wiener space method, a high-order weak approximation scheme,\nis established for SPDEs in the case of unbounded characteristics and unbounded\npayoffs. We first introduce a recently described flexible functional analytic\nframework, so called weighted spaces, where Feller-like properties hold. A\nrefined analysis of vector fields on weighted spaces then yields optimal\nconvergence rates of cubature methods for stochastic partial differential\nequations of Da Prato-Zabczyk type. The ubiquitous stability for the local\napproximation operator within the functional analytic setting is proved for\nSPDEs, however, in the infinite dimensional case we need a newly introduced\nassumption on weak symmetry of the cubature formula. In finite dimensions, we\nuse the UFG condition to obtain optimal rates of convergence on non-uniform\nmeshes for nonsmooth payoffs with exponential growth.\n", " Discrete geometric estimators aim at estimating geometric characteristics of\na shape with only its digitization as input data. Such an estimator is\nmultigrid convergent when its estimates tend toward the geometric\ncharacteristics of the shape as the digitization step h tends toward 0. This\npaper studies the multigrid convergence of tangent estimators based on maximal\ndigital straight segment recognition. We show that such estimators are\nmultigrid convergent for some family of convex shapes and that their speed of\nconvergence is on average O(h^(2/3)). Experiments confirm this result and\nsuggest that the bound is tight.\n", " A method of proving local continuity of concave functions on convex set\npossessing the $\\mu$-compactness property is presented. This method is based on\na special approximation of these functions.\n The class of $\\mu$-compact sets can be considered as a natural extension of\nthe class of compact metrizable subsets of locally convex spaces, to which\nparticular results well known for compact sets can be generalized.\n Applications of the obtained continuity conditions to analysis of different\nentropic characteristics of quantum systems and channels are considered.\n", " On the space of bounded analytic functions and the Bloch space on the unit\ndisk, we study the compact intertwining relations for composition operators,\nwhose intertwining operators are Volterra type operators. Further, we consider\nthe compact intertwining relations, which are between the whole collection of\ncomposition operators and some Volterra operator, and the whole collection of\nbounded Volterra operators and some composition operator.\n" ]

[ " Both algebras, Clifford and Grassmann, offer \"basis vectors\" for describing\nthe internal degrees of freedom of fermions. The oddness of the \"basis\nvectors\", transferred to the creation operators, which are tensor products of\nthe finite number of \"basis vectors\" and the infinite number of momentum basis,\nand to their Hermitian conjugated partners annihilation operators, offers the\nsecond quantization of fermions without postulating the conditions proposed by\nDirac, enabling the explanation of the Dirac's postulates. But while the\nClifford fermions manifest the half integer spins -- in agreement with the\nobserved properties of quarks and leptons and antiquarks and antileptons -- the\n\"Grassmann fermions\" manifest the integer spins. In Part I properties of the\ncreation and annihilation operators of integer spins \"Grassmann fermions\" are\npresented and the proposed equations of motion solved. The anticommutation\nrelations of second quantized integer spin fermions are shown when applying on\nthe vacuum state as well as when applying on the Hilbert space of the infinite\nnumber of \"Slater determinants\" with all the possibilities of empty and\noccupied \"fermion states\". In Part II the conditions are discussed under which\nthe Clifford algebras offer the appearance of the second quantized fermions,\nenabling as well the appearance of families. In both parts, Part I and Part II,\nthe relation between the Dirac way and our way of the second quantization of\nfermions is presented.\n", " Given a complex Hilbert space H and the von Neumann algebra L(H) of all\nbounded linear operators on H, we study the Grassmann manifold M of all\nprojections in L(H) that have a fixed finite rank r. We take the Jordan-Banach\ntriple theory approach which allows us to define a natural Levi-Civita\nconnection on M. We identify the geodesics, compute the Riemann distance and\nprove some properties of M\n", " The hyperbolic algebra A_h, studied recently by Katavolos and Power, is the\nweak star closed operator algebra on L^2(R) generated by H^\\infty(R), as\nmultiplication operators, and by the dilation operators V_t, t \\geq 0, given by\nV_t f(x) = e^{t/2} f(e^t x). We show that A_h is a reflexive operator algebra\nand that the four dimensional manifold Lat A_h (with the natural topology) is\nthe reflexive hull of a natural two dimensional subspace.\n", " It is shown that (special) relativistically dilated time is the vector sum of\nrest time and time induced by movement in three dimensional space exeeding the\nrest time component and that the first vector is orthogonally directed relative\nto our three dimensional space. This again implies that its origin lies in the\nmovement of three dimensional space relative to a four dimensional manifold.\nThe theory predicts asymmetrical time dilation.\n", " We develop a geometric invariant Littlewood-Paley theory for arbitrary\ntensors on a compact 2 dimensional manifold. We show that all the important\nfeatures of the classical LP theory survive with estimates which depend only on\nvery limited regularity assumptions on the metric. We give invariant\ndescriptions of Sobolev and Besov spaces and prove some sharp product\ninequalities. This theory has being developed in connection to the work of the\nauthors on the geometry of null hypersurfaces with a finite curvature flux\ncondition, see \\cite{KR1}, \\cite{KR3}.\n We are confident however that it can be applied, and extended, to many\ndifferent situations.\n", " In this paper we study stable ergodicity of the action of groups of\ndiffeomorphisms on smooth manifolds. The existence of such actions is known\nonly on one dimensional manifolds. The aim of this paper is to introduce a\ngeometric method to overcome this restriction and for constructing higher\ndimensional examples. In particular, we show that every closed manifold admits\nstably ergodic finitely generated group actions by diffeomorphisms of class\n$C^{1+\\alpha}$. We also prove the stable ergodicity of certain algebraic\nactions including the natural action of a generic pair of matrices near the\nidentity on a sphere of arbitrary dimension. These are consequences of the\nquasi-conformal blender, a local and stable mechanism/phenomenon introduced in\nthis paper, which encapsulates our method to prove stable local ergodicity by\nproviding quasi-conformal orbits with fine controlled geometry. The\nquasi-conformal blender is developed in the context of pseudo-semigroup actions\nof locally defined smooth diffeomorphisms. This allows for applications in\ndifferent settings, including for the smooth foliations of arbitrary\ncodimension.\n", " Surprisingly, the issue of events localization in spacetime is poorly\nunderstood and a fortiori realized even in the context of Einstein's\nrelativity. Accordingly, a comparison between observational data and\ntheoretical expectations might then be strongly compromised. In the present\npaper, we give the principles of relativistic localizing systems so as to\nbypass this issue. Such systems will allow to locate users endowed with\nreceivers and, in addition, to localize any spacetime event. These localizing\nsystems are made up of relativistic auto-locating positioning sub-systems\nsupplemented by an extra satellite. They indicate that spacetime must be\nsupplied everywhere with an unexpected local four dimensional projective\nstructure besides the well-known three dimensional relativistic projective one.\nAs a result, the spacetime manifold can be seen as a generalized Cartan space\nmodeled on a four dimensional real projective space, i.e., a spacetime with\nboth a local four dimensional projective structure and a compatible\n(pseudo-)Riemannian structure. Localization protocols are presented in details,\nwhile possible applications to astrophysics are also considered.\n", " Using symplectic techniques and spectral analysis of smooth paths of\nself-adjoint operators, we characterize the set of conjugate instants along a\ngeodesic in an infinite dimensional Riemannian Hilbert manifold.\n", " This paper introduces the concept of functional current as a mathematical\nframework to represent and treat functional shapes, i.e. sub-manifold supported\nsignals. It is motivated by the growing occurrence, in medical imaging and\ncomputational anatomy, of what can be described as geometrico-functional data,\nthat is a data structure that involves a deformable shape (roughly a finite\ndimensional sub manifold) together with a function defined on this shape taking\nvalue in another manifold.\n Indeed, if mathematical currents have already proved to be very efficient\ntheoretically and numerically to model and process shapes as curves or\nsurfaces, they are limited to the manipulation of purely geometrical objects.\nWe show that the introduction of the concept of functional currents offers a\ngenuine solution to the simultaneous processing of the geometric and signal\ninformation of any functional shape. We explain how functional currents can be\nequipped with a Hilbertian norm mixing geometrical and functional content of\nfunctional shapes nicely behaving under geometrical and functional\nperturbations and paving the way to various processing algorithms. We\nillustrate this potential on two problems: the redundancy reduction of\nfunctional shapes representations through matching pursuit schemes on\nfunctional currents and the simultaneous geometric and functional registration\nof functional shapes under diffeomorphic transport.\n", " We study the Helton-Howe trace and the Connes-Chern character for Toeplitz\noperators on weighted Bergman spaces via the idea of quantization. We prove a\nlocal formula for the large $t$-limit of the Connes-Chern character as the\nweight goes to infinity. And we show that the Helton-Howe trace of Toeplitz\noperators is independent of the weight $t$ and obtain a local formula for the\nHelton-Howe trace for all weighted Bergman spaces using harmonic analysis and\nquantization.\n", " We give a functional analytic construction of the fermionic projector on a\nglobally hyperbolic Lorentzian manifold of finite lifetime. The integral kernel\nof the fermionic projector is represented by a two-point distribution on the\nmanifold. By introducing an ultraviolet regularization, we get to the framework\nof causal fermion systems. The connection to the \"negative-energy solutions\" of\nthe Dirac equation and to the WKB approximation is explained and quantified by\na detailed analysis of closed Friedmann-Robertson-Walker universes.\n", " We show that edge states similar to those known for topological insulators\nexist in two-dimensional electron system with one-band spectrum in the presence\nof heterogeneous spin-orbit interaction (SOI). These states appear at\nboundaries between regions with the SOIs of different kind or between the\nregions with the SOI and without it. Depending on the system parameters they\ncan appear in an energy range lying both in the forbidden and conduction bands\nof bulk states. The edge states have chiral spin texture and carry a spin\ncurrent under the equilibrium. We study also the size quantization of the edge\nstates in a strip structure with two boundaries to find an unusual dependence\nof the quantization energy on the strip width.\n", " We focus on the problem of manifold estimation: given a set of observations\nsampled close to some unknown submanifold $M$, one wants to recover information\nabout the geometry of $M$. Minimax estimators which have been proposed so far\nall depend crucially on the a priori knowledge of some parameters quantifying\nthe underlying distribution generating the sample (such as bounds on its\ndensity), whereas those quantities will be unknown in practice. Our\ncontribution to the matter is twofold: first, we introduce a one-parameter\nfamily of manifold estimators $(\\hat{M}_t)_{t\\geq 0}$ based on a localized\nversion of convex hulls, and show that for some choice of $t$, the\ncorresponding estimator is minimax on the class of models of $C^2$ manifolds\nintroduced in [Genovese et al., Manifold estimation and singular deconvolution\nunder Hausdorff loss]. Second, we propose a completely data-driven selection\nprocedure for the parameter $t$, leading to a minimax adaptive manifold\nestimator on this class of models. This selection procedure actually allows us\nto recover the Hausdorff distance between the set of observations and $M$, and\ncan therefore be used as a scale parameter in other settings, such as tangent\nspace estimation.\n", " This paper identifies certain interesting mathematical problems of stochastic\nquantization type in the modeling of Laser propagation through turbulent media.\nIn some of the typical physical contexts the problem reduces to stochastic\nSchrodinger equation with space-time white noise of Gaussian, Poisson and Levy\ntype. We identify their mathematical resolution via stochastic quantization.\nNonlinear phenomena such as Kerr effect can be modeled by stochastic nonlinear\nSchrodinger equation in the focusing case with space-time white noise. A\ntreatment of stochastic transport equation, the Korteweg-de Vries Equation as\nwell as a number of other nonlinear wave equations with space-time white noise\nis also given. Main technique is the S-transform (we will actually use closely\nrelated Hermite transform) which converts the stochastic partial differential\nequation with space time white noise to a deterministic partial differential\nequation defined on the Hida-Kondratiev white noise distribution space. We then\nutlize the inverse S-transform/Hermite transform known as the characterization\ntheorem combined with the infinite dimensional implicit function theorem for\nanalytic maps to establish local existence and uniqueness theorems for pathwise\nsolutions of these class of problems. The particular focus of this paper on\nsingular white noise distributions is motivated by practical situations where\nthe refractive index fluctuations in propagation medium in space and time are\nintense due to turbulence, ionospheric plasma turbulence, marine-layer\nfluctuations, etc. Since a large class of partial differential equations that\narise in nonlinear wave propagation have polynomial type nonlinearities, white\nnoise distribution theory is an effective tool in studying these problems\nsubject to different types of white noises.\n", " Hydrodynamic attractors have recently gained prominence in the context of\nearly stages of ultra-relativistic heavy-ion collisions at the RHIC and LHC. We\ncritically examine the existing ideas on this subject from a phase space point\nof view. In this picture the hydrodynamic attractor can be seen as a special\ncase of the more general phenomenon of dynamical dimensionality reduction of\nphase space regions. We quantify this using Principal Component Analysis.\nFurthermore, we adapt the well known slow-roll approximation to this setting.\nThese techniques generalize easily to higher dimensional phase spaces, which we\nillustrate by a preliminary analysis of a dataset describing the evolution of a\n5-dimensional manifold of initial conditions immersed in a 16-dimensional\nrepresentation of the phase space of the Boltzmann kinetic equation in the\nrelaxation time approximation.\n", " Rabern recently proved that any graph with omega >= (3/4)(Delta+1) contains a\nstable set meeting all maximum cliques. We strengthen this result, proving that\nsuch a stable set exists for any graph with omega > (2/3)(Delta+1). This is\ntight, i.e. the inequality in the statement must be strict. The proof relies on\nfinding an independent transversal in a graph partitioned into vertex sets of\nunequal size.\n", " Based on the electromagnetic interaction and gravitational interaction, the\nquantum interplays of the ordinary matter and the dark matter in the octonion\nspace are discussed. The paper presents the quantization of the particles of\nordinary matter and dark matter, including the quantization of the\nelectromagnetic field and gravitational field etc. In the electromagnetic and\ngravitational interactions, it deduces some predictions of the field source\nparticle, which are consistent with the Dirac equation and Schrodinger equation\nin the quantum mechanics. The researches show that there exist some quantum\ncharacteristics of the intermediate particle in the electromagnetic and\ngravitational interactions, including the Dirac-like equation etc.\n", " Smooth manifolds are not the suitable context for trying to generalize the\nconcept of rough paths on a manifold. Indeed, when one is working with smooth\nmaps instead of Lipschitz maps and trying to solve a rough differential\nequation, one loses the quantitative estimates controlling the convergence of\nthe Picard sequence. Moreover, even with a definition of rough paths in smooth\nmanifolds, ordinary and rough differential equations can only be solved locally\nin such case. In this paper, we first recall the foundations of the Lipschitz\ngeometry, introduced in \"Rough Paths on Manifolds\" (Cass, T., Litterer, C. &\nLyons, T.), along with the main findings that encompass the classical theory of\nrough paths in Banach spaces. Then we give what we believe to be a minimal\nframework for defining rough paths on a manifold that is both less rigid than\nthe classical one and emphasized on the local behaviour of rough paths. We end\nby explaining how this same idea can be used to define any notion of coloured\npaths on a manifold.\n", " We use Klee's Dehn-Sommerville relations and other results on face numbers of\nhomology manifolds without boundary to (i) prove Kalai's conjecture providing\nlower bounds on the f-vectors of an even-dimensional manifold with all but the\nmiddle Betti number vanishing, (ii) verify K\\\"uhnel's conjecture that gives an\nupper bound on the middle Betti number of a 2k-dimensional manifold in terms of\nk and the number of vertices, and (iii) partially prove K\\\"uhnel's conjecture\nproviding upper bounds on other Betti numbers of odd- and even-dimensional\nmanifolds. For manifolds with boundary, we derive an extension of Klee's\nDehn-Sommerville relations and strengthen Kalai's result on the number of their\nedges.\n", " An inconsistency of quantum field theory, regarding the signs of vacuum\nenergy and vacuum pressure of elementary fields versus non-elementary fields\n(like e.g. phonon fields), is pointed out. An improved law for the canonical\nquantization of fields is presented, which is based on the distinction between\nelementary and non-elementary fields. Remarkably, the improved quantization\nmethod removes not only the inconsistency of quantum field theory, but at the\nsame time solves the (old) cosmological constant problem for all fields of the\nstandard model of elementary particles (but not for the hypothetical inflaton\nfields), without compromising any of the achievements of established quantum\nfield theory.\n", " In this paper, we proved the correspondence between Feynman diagrams in\nspace-time and light-cone diagrams in world-sheet by using only path integral\nrepresentation on free Green function in the first quantization theory. We also\nobtained general representation on perturbation series of light-cone diagrams\ndescribing split and join of strings.\n", " Boolean, free and monotone cumulants as well as relations among them, have\nproven to be important in the study of non-commutative probability theory.\nQuite notably, Boolean cumulants were successfully used to study free infinite\ndivisibility via the Boolean Bercovici--Pata bijection. On the other hand, in\nrecent years the concept of infinitesimal non-commutative probability has been\ndeveloped, together with the notion of infinitesimal cumulants which can be\nuseful in the context of combinatorial questions.\n In this paper, we show that the known relations among free, Boolean and\nmonotone cumulants still hold in the infinitesimal framework. Our approach is\nbased on the use of Grassmann algebra. Formulas involving infinitesimal\ncumulants can be obtained by applying a formal derivation to known formulas.\n The relations between the various types of cumulants turn out to be captured\nvia the shuffle algebra approach to moment-cumulant relations in\nnon-commutative probability theory. In this formulation, (free, Boolean and\nmonotone) cumulants are represented as elements of the Lie algebra of\ninfinitesimal characters over a particular combinatorial Hopf algebra. The\nlatter consists of the graded connected double tensor algebra defined over a\nnon-commutative probability space and is neither commutative nor cocommutative.\nIn this note it is shown how the shuffle algebra approach naturally extends to\nthe notion of infinitesimal non-commutative probability space. The basic step\nconsists in replacing the base field as target space of linear Hopf algebra\nmaps by the Grassmann algebra over the base field. We also consider the\ninfinitesimal analog of the Boolean Bercovici--Pata map.\n", " We show that submultiplicative norms on section rings of polarised projective\nmanifolds are asymptotically equivalent to sup-norms associated with metrics on\nthe polarisation.\n As an application, we establish that over canonically polarised manifolds,\nconvex hull of Narasimhan-Simha pseudonorm over pluricanonical sections is\nasymptotically equivalent to the sup-norm associated with the supercanonical\nmetric of Tsuji, refining a result of Berman-Demailly.\n As another application, we deduce that the jumping measures associated with\nbounded submultiplicative filtrations on section rings converge to the spectral\nmeasures of their Bergman geodesic rays, generalizing previous results of Witt\nNystr\\\"om and Hisamoto. We show also that the latter measures can be described\nusing pluripotential theory. More precisely, we establish that Bergman geodesic\nrays are maximal, i.e. they can be constructed through plurisubharmonic\nenvelopes.\n As a final application, for non-continuous metrics on ample line bundles over\nprojective manifolds, we prove that a weak version of semiclassical holomorphic\nextension theorem holds for generic submanifolds. This means that up to a\nnegligible portion, the totality of holomorphic sections over generic\nsubmanifolds extends in an effective way to the ambient manifold.\n As an unexpected byproduct, we show that injective and projective tensor\nnorms on symmetric algebras of finitely dimensional complex normed vector\nspaces are asymptotically equivalent.\n", " We examine in detail the changes in the production of primordial helium\nresulting from nonthermal neutrino momentum distributions produced by resonant\ntransformation of electron-type neutrinos to steriles. These transformations,\nanti-nu_e -> anti-nu_s (nu_e->nu_s), amplify a positive (negative) lepton\nnumber asymmetry. We find that the resulting supression relative to a thermal\ndistribution of low energy nu_e reduces n -> p conversion to a greater extent\nthan does the enhancement of n -> p from an identical suppression of anti-nu_e.\nThus, equal lepton-number asymmetries of opposite sign have unequal effects on\nthe resulting helium yield in primordial nucleosynthesis.\n", " We consider the tree amplitudes of production of $n_2$ scalar particles by\n$n_1$ particles of another kind, where both initial and final particles are at\nrest and on mass shell, in a model of two scalar fields with $O(2)$ symmetric\ninteraction and unequal masses. We find that these amplitudes are zero except\nfor the lowest possible $n_1$ and $n_2$, and that the cancellation of the\ncorresponding Feynman graphs occurs due to a special symmetry of the classical\nmechanical counterpart of this theory. This feature is rather general and is\ninherent in various other scalar field theories.\n", " This is the lecture notes on the interplay between optimal transport and\nRiemannian geometry. On a Riemannian manifold, the convexity of entropy along\noptimal transport in the space of probability measures characterizes lower\nbounds of the Ricci curvature. We then discuss geometric properties of general\nmetric measure spaces satisfying this convexity condition.\n", " In this article, we define the spherical $\\pi$-operator over domains in the\n$(n-1)$-D unit sphere $S^n$ of $R^n$ and develop new and analogous results on\nthe operator it self and its mapping properties. We introduce the spherical\nDirac operator $\\Gamma_\\alpha$ as an $\\alpha$- shift of of $\\Gamma_omega$,\nwhere $\\Gamma_omega$ is the negative of the wedge (or Grassmann) product of\n$\\omega$ with that of the Dirac operator $D_\\omega$. A gegenbauer polynomial is\nused as a Cauchy kernel for the spherical Dirac operator $\\Gamma_alpha$.\n", " In this article we extend the computational geometric curve reconstruction\napproach to curves in Riemannian manifolds. We prove that the minimal spanning\ntree, given a sufficiently dense sample, correctly reconstructs the smooth arcs\nand further closed and simple curves in Riemannian manifolds. The proof is\nbased on the behaviour of the curve segment inside the tubular neighbourhood of\nthe curve. To take care of the local topological changes of the manifold, the\ntubular neighbourhood is constructed in consideration with the injectivity\nradius of the underlying Riemannian manifold. We also present examples of\nsuccessfully reconstructed curves and show an applications of curve\nreconstruction to ordering motion frames.\n", " A while ago a proposal have been made regarding Klein Gordon and Maxwell\nLagrangians for causal set theory. These Lagrangian densities are based on the\nstatistical analysis of the behavior of field on a sample of points taken\nthroughout some \"small\" region of spacetime. However, in order for that sample\nto be statistically reliable, a lower bound on the size of that region needs to\nbe imposed. This results in \"unwanted contributions\" from higher order\nderivatives to the Lagrangian density, as well as non-trivial curvature effects\non the latter. It turns out that both gravitational and non-gravitational\neffects end up being highly non-linear. In the previous papers we were focused\non leading order terms, which allowed us to neglect these nonlinearities. We\nwould now like to go to the next order and investigate them. In the current\npaper we will exclusively focus on the effects of higher order derivatives in\nthe flat-space toy model. The gravitational effects will be studied in another\npaper which is currently in preparation. Both papers are restricted to bosonic\nfields, although the issue probably generalizes to fermions once Grassmann\nnumbers are dealt with in appropriate manner.\n", " In 1990 van Eijnghoven and Meyers introduced systems of holomorphic Hermite\nfunctions and reproducing kernel Hilbert spaces associated with the systems on\nthe complex plane. Moreover they studied the relationship between the family of\nall their Hilbert spaces and a class of Gelfand-Shilov functions. After that,\ntheir systems of holomorphic Hermite functions have been applied to studying\nquantization on the complex plane, combinatorics, and etc. On the other hand,\nthe author recently introduced systems of holomorphic Hermite functions\nassociated with ellipses on the complex plane. The present paper shows that\ntheir systems of holomorphic Hermite functions are determined by some cases of\nellipses, and that their reproducing kernel Hilbert spaces are some cases of\nthe Segal-Bargmann spaces determined by the Bargmann-type transforms introduced\nby Sjoestrand.\n", " We prove an analogue of Bonami's (hypercontractive) lemma for complex-valued\nfunctions on $\\mathcal{L}(V,W)$, where $V$ and $W$ are vector spaces over a\nfinite field. This inequality is useful for functions on $\\mathcal{L}(V,W)$\nwhose `generalised influences' are small, in an appropriate sense. It leads to\na significant shortening of the proof of a recent seminal result by Khot,\nMinzer and Safra that pseudorandom sets in Grassmann graphs have near-perfect\nexpansion, which (in combination with the work of Dinur, Khot, Kindler, Minzer\nand Safra) implies the 2-2 Games conjecture (the variant, that is, with\nimperfect completeness).\n", " Consider the following distributed optimization scenario. A worker has access\nto training data that it uses to compute the gradients while a server decides\nwhen to stop iterative computation based on its target accuracy or delay\nconstraints. The server receives all its information about the problem instance\nfrom the worker via a rate-limited noiseless communication channel. We\nintroduce the principle we call Differential Quantization (DQ) that prescribes\ncompensating the past quantization errors to direct the descent trajectory of a\nquantized algorithm towards that of its unquantized counterpart. Assuming that\nthe objective function is smooth and strongly convex, we prove that\nDifferentially Quantized Gradient Descent (DQ-GD) attains a linear contraction\nfactor of $\\max\\{\\sigma_{\\mathrm{GD}}, \\rho_n 2^{-R}\\}$, where\n$\\sigma_{\\mathrm{GD}}$ is the contraction factor of unquantized gradient\ndescent (GD), $\\rho_n \\geq 1$ is the covering efficiency of the quantizer, and\n$R$ is the bitrate per problem dimension $n$. Thus at any $R\\geq\\log_2 \\rho_n\n/\\sigma_{\\mathrm{GD}}$ bits, the contraction factor of DQ-GD is the same as\nthat of unquantized GD, i.e., there is no loss due to quantization. We show\nthat no algorithm within a certain class can converge faster than\n$\\max\\{\\sigma_{\\mathrm{GD}}, 2^{-R}\\}$. Since quantizers exist with $\\rho_n \\to\n1$ as $n \\to \\infty$ (Rogers, 1963), this means that DQ-GD is asymptotically\noptimal. The principle of differential quantization continues to apply to\ngradient methods with momentum such as Nesterov's accelerated gradient descent,\nand Polyak's heavy ball method. For these algorithms as well, if the rate is\nabove a certain threshold, there is no loss in contraction factor obtained by\nthe differentially quantized algorithm compared to its unquantized counterpart.\nExperimental results on least-squares problems validate our theoretical\nanalysis.\n", " The aim of this paper is to prove a stronger version of a conjecture on the\nexistence of non-dominated scalar-valued m-homogeneous polynomials (m>=3) on\narbitrary infinite dimensional Banach spaces.\n", " We present a new method based on functional tensor decomposition and dynamic\ntensor approximation to compute the solution of a high-dimensional\ntime-dependent nonlinear partial differential equation (PDE). The idea of\ndynamic approximation is to project the time derivative of the PDE solution\nonto the tangent space of a low-rank functional tensor manifold at each time.\nSuch a projection can be computed by minimizing a convex energy functional over\nthe tangent space. This minimization problem yields the unique optimal velocity\nvector that allows us to integrate the PDE forward in time on a tensor manifold\nof constant rank. In the case of initial/boundary value problems defined in\nreal separable Hilbert spaces, this procedure yields evolution equations for\nthe tensor modes in the form of a coupled system of one-dimensional\ntime-dependent PDEs. We apply the dynamic tensor approximation to a\nfour-dimensional Fokker-Planck equation with non-constant drift and diffusion\ncoefficients, and demonstrate its accuracy in predicting relaxation to\nstatistical equilibrium.\n", " Bandstructure effects in the electronic transport of strongly quantized\nsilicon nanowire field-effect-transistors (FET) in various transport\norientations are examined. A 10-band sp3d5s* semi-empirical atomistic\ntight-binding model coupled to a self consistent Poisson solver is used for the\ndispersion calculation. A semi-classical, ballistic FET model is used to\nevaluate the current-voltage characteristics. It is found that the total gate\ncapacitance is degraded from the oxide capacitance value by 30% for wires in\nall the considered transport orientations ([100], [110], [111]). Different wire\ndirections primarily influence the carrier velocities, which mainly determine\nthe relative performance differences, while the total charge difference is\nweakly affected. The velocities depend on the effective mass and degeneracy of\nthe dispersions. The [110] and secondly the [100] oriented 3nm thick nanowires\nexamined, indicate the best ON-current performance compared to [111] wires. The\ndispersion features are strong functions of quantization. Effects such as\nvalley splitting can lift the degeneracies especially for wires with cross\nsection sides below 3nm. The effective masses also change significantly with\nquantization, and change differently for different transport orientations. For\nthe cases of [100] and [111] wires the masses increase with quantization,\nhowever, in the [110] case, the mass decreases. The mass variations can be\nexplained from the non-parabolicities and anisotropies that reside in the first\nBrillouin zone of silicon.\n", " Using the concept of projective systems for linear codes and elementary\nlinear algebra, we show that projective $[n,k]_q$ codes form a connected\nsubgraph in the Grassmann graph consisting of $k$-dimensional subspaces of an\n$n$-dimensional vector space over the $q$-element field.\n", " For the wave equation associated to the Laplacian on a compact manifold with\nboundary with a conic metric (with respect to which the boundary is metrically\na point) the propagation of singularities through the boundary is analyzed.\nUnder appropriate regularity assumptions the diffracted, non-direct, wave\nproduced by the boundary is shown to have Sobolev regularity greater than the\nincoming wave.\n", " Often noisy point clouds are given as an approximation of a particular\ncompact set of interest. A finite point cloud is a compact set. This paper\nproves a reconstruction theorem which gives a sufficient condition, as a bound\non the Hausdorff distance between two compact sets, for when certain offsets of\nthese two sets are homotopic in terms of the absence of {\\mu}-critical points\nin an annular region. Since an offset of a set deformation retracts to the set\nitself provided that there are no critical points of the distance function\nnearby, we can use this theorem to show when the offset of a point cloud is\nhomotopy equivalent to the set it is sampled from. The ambient space can be any\nRiemannian manifold but we focus on ambient manifolds which have nowhere\nnegative curvature. In the process, we prove stability theorems for\n{\\mu}-critical points when the ambient space is a manifold.\n", " We define operator manifolds as manifolds on which a spectral measure on a\nHilbert space is given as additional structure. The spectral measure\nmathematically describes space as a quantum mechanical observable.\n We show that the vectors of the Hilbert space can be represented as functions\non the manifold. The arbitrariness of this representation is interpreted as\nlocal gauge freedom. In this way, the physical gauge principle is linked with\nquantum mechanical measurements of the position variable. We derive the\nrestriction for the local gauge group to be U(m), where m is the number of\ncomponents of the wave functions.\n", " In this paper the Landscape potential is considered as an environment for\nsystem: trajectory-environment (TE). The trajectory is generating a measure on\nthe landscape. The entropy of this dynamic measure is a power factor for\ntrajectory. This dynamics leads to a memory appearance and produces a\nnon-singular measure practically independently from initial conditions. It is\nshown that measure is dual to the metrics and its evolution may be considered\nas the landscape deformations: production-destruction of attractors-vacua. It\nseems like the renormalization process with phase transitions: renormgroup\nquantization. It is shown what the entropy of this global landscape measure\nacts on the trajectory alike a dark energy.\n", " We introduce quantum spatio-temporal dynamics (QSD) as modeled by the Nexus\nParadigm (NP) of quantum gravity to resolve the problem of energy-momentum\nlocalization in a gravitational field. Currently, the gravitational field as\ndescribed using the language of geometry modeled under General Relativity (GR)\nfails to provide a generally accepted definition of energy-momentum. Attempts\nat resolving this problem using geometric methods have resulted in various\nenergy-momentum complexes whose physical meaning remain dubious since the\nresulting complexes are non-tensorial under a general coordinate\ntransformation. In QSD, the tangential manifold is the affine connection field\nin which energy-momentum localization is readily defined. We also discover that\nthe positive mass condition is a natural consequence of quantization and that\ndark energy is a Higgs like field with negative energy density everywhere.\nFinally, energy-momentum localization in quantum gravity shows that a free\nfalling object will experience larger vacuum fluctuations (uncertainties in\nlocation) in strong gravity than in weak gravity and that the amplitudes of\nthese oscillations define the energy of the free falling object.\n", " Given a sample of an abstract manifold immersed in some Euclidean space, we\ndescribe a way to recover the singular homology of the original manifold. It\nconsists in estimating its tangent bundle -- seen as subset of another\nEuclidean space -- in a measure theoretic point of view, and in applying\nmeasure-based filtrations for persistent homology. The construction we propose\nis consistent and stable, and does not involve the knowledge of the dimension\nof the manifold. In order to obtain quantitative results, we introduce the\nnormal reach, which is a notion of reach suitable for an immersed manifold.\n", " A human does not have to see all elephants to recognize an animal as an\nelephant. On contrast, current state-of-the-art deep learning approaches\nheavily depend on the variety of training samples and the capacity of the\nnetwork. In practice, the size of network is always limited and it is\nimpossible to access all the data samples. Under this circumstance, deep\nlearning models are extremely fragile to human-imperceivable adversarial\nexamples, which impose threats to all safety critical systems. Inspired by the\nassociation and attention mechanisms of the human brain, we propose reverse\nadversarial examples method that can greatly improve models' robustness on\nunseen data. Experiments show that our reverse adversarial method can improve\naccuracy on average 19.02% on ResNet18, MobileNet, and VGG16 on unseen data\ntransformation. Besides, the proposed method is also applicable to compressed\nmodels and shows potential to compensate the robustness drop brought by model\nquantization - an absolute 30.78% accuracy improvement.\n", " There must exist a reformulation of quantum field theory, even at low\nenergies, which does not depend on classical time. The octonionic theory\nproposes such a reformulation, leading to a pre-quantum pre-spacetime theory.\nThe ingredients for constructing such a theory, which is also a unification of\nthe standard model with gravitation, are : (i) the pre-quantum theory of trace\ndynamics: a matrix-valued Lagrangian dynamics, (ii) the spectral action\nprinciple of non-commutative geometry, (iii) the number system known as the\noctonions, for constructing a non-commutative manifold and for defining\nelementary particles via Clifford algebras, (iv) a Lagrangian with $E_8 \\times\nE_8$ symmetry. The split bioctonions define a sixteen dimensional space (with\nleft-right symmetry) whose geometry (evolving in Connes time) relates to the\nfour known fundamental forces, while predicting two new forces, $SU(3)_{grav}$\nand $U(1)_{grav}$. This latter interaction is possibly the theoretical origin\nof MOND. Coupling constants of the standard model result from left-right\nsymmetry breaking, and their values are theoretically determined by the\ncharacteristic equation of the exceptional Jordan algebra of the octonions. The\nquantum-to-classical transition, precipitated by the entanglement of a critical\nnumber of fermions, is responsible for the emergence of classical spacetime,\nand also for the familiar formulation of quantum theory on a spacetime\nbackground.\n", " We introduce an intrinsic estimator for the scalar curvature of a data set\npresented as a finite metric space. Our estimator depends only on the metric\nstructure of the data and not on an embedding in $\\mathbb{R}^n$. We show that\nthe estimator is consistent in the sense that for points sampled from a\nprobability measure on a compact Riemannian manifold, the estimator converges\nto the scalar curvature as the number of points increases. To justify its use\nin applications, we show that the estimator is stable with respect to\nperturbations of the metric structure, e.g., noise in the sample or error\nestimating the intrinsic metric. We validate our estimator experimentally on\nsynthetic data that is sampled from manifolds with specified curvature.\n", " Let $\\mathscr{L}$ be a smooth second-order real differential operator in\ndivergence form on a manifold of dimension $n$. Under a bracket-generating\ncondition, we show that the ranges of validity of spectral multiplier estimates\nof Mihlin--H\\\"ormander type and wave propagator estimates of Miyachi--Peral\ntype for $\\mathscr{L}$ cannot be wider than the corresponding ranges for the\nLaplace operator on $\\mathbb{R}^n$. The result applies to all sub-Laplacians on\nCarnot groups and more general sub-Riemannian manifolds, without restrictions\non the step. The proof hinges on a Fourier integral representation for the wave\npropagator associated with $\\mathscr{L}$ and nondegeneracy properties of the\nsub-Riemannian geodesic flow.\n", " In this paper, we will first clarify the physical meaning of having a minimum\nmeasurable time. Then we will combine the deformation of the Dirac equation due\nto the existence of minimum measurable length and time scales with its\ndeformation due to the doubly special relativity. We will also analyse this\ndeformed Dirac equation in curved spacetime, and observe that this deformation\nof the Dirac equation also leads to a non-trivial modification of general\nrelativity. Finally, we will analyse the stochastic quantization of this\ndeformed Dirac equation on curved spacetime.\n", " These notes give an introduction to the Gabriel-Roiter measure of a finite\ndimensional algebra. They are based on a series of four lectures at the\n\"Advanced School and Conference on Representation Theory and Related Topics\" in\nTrieste (ICTP, January 2006).\n", " For a class of competitive maps there is an invariant one-codimensional\nmanifold (the carrying simplex) attracting all non-trivial orbits. In the\npresent paper it is shown that its convexity implies that it is a $C^1$\nsubmanifold-with-corners, neatly embedded in the non-negative orthant. The\nproof uses the characterization of neat embedding in terms of inequalities\nbetween Lyapunov exponents for ergodic invariant measures supported on the\nboundary of the carrying simplex.\n", " The necessity of renormalization arises from the infinite integrals which are\ncaused by the discrepancy between the orders of differential and integral\noperators in the four dimensional QFTs. Therefore in view of the fact that\nfiniteness and invariant properties of operators are their topological aspects,\nessential renormalization tools to extract finite invariant values from those\ninfinities which are comparable with the experimental results, e. g.\nregularization, perturbation and radiative corrections follow some topological\nstandards. In the second part we consider dimensional and geometrical\nconsequences of topological approach to renormalization for the geometrical\nstructure and degrees of freedom of renormalized theory. We show that\nregularization and renormalization of QED are performed only by certain\nrestrictive dimensional conditions on QED fields. Further it is shown that in\naccord with our previous topological approach to renormalization of QED the\ngeometrical evaluation of applied dimensional renormalization conditions and\nthe appearance of anomalies refer to a reduction of number of degrees of\nfreedom according to the reduced symmetry of QED. A conclusion concerning a\ncomparison of our results with holographic principle models is also included.\n" ]

[ " Trajectory prediction is critical for autonomous driving vehicles. Most\nexisting methods tend to model the correlation between history trajectory\n(input) and future trajectory (output). Since correlation is just a superficial\ndescription of reality, these methods rely heavily on the i.i.d. assumption and\nevince a heightened susceptibility to out-of-distribution data. To address this\nproblem, we propose an Out-of- Distribution Causal Graph (OOD-CG), which\nexplicitly defines the underlying causal structure of the data with three\nentangled latent features: 1) domain-invariant causal feature (IC), 2)\ndomain-variant causal feature (VC), and 3) domain-variant non-causal feature\n(VN ). While these features are confounded by confounder (C) and domain\nselector (D). To leverage causal features for prediction, we propose a Causal\nInspired Learning Framework (CILF), which includes three steps: 1) extracting\ndomain-invariant causal feature by means of an invariance loss, 2) extracting\ndomain variant feature by domain contrastive learning, and 3) separating\ndomain-variant causal and non-causal feature by encouraging causal sufficiency.\nWe evaluate the performance of CILF in different vehicle trajectory prediction\nmodels on the mainstream datasets NGSIM and INTERACTION. Experiments show\npromising improvements in CILF on domain generalization.\n", " Estimating causal relations is vital in understanding the complex\ninteractions in multivariate time series. Non-linear coupling of variables is\none of the major challenges inaccurate estimation of cause-effect relations. In\nthis paper, we propose to use deep autoregressive networks (DeepAR) in tandem\nwith counterfactual analysis to infer nonlinear causal relations in\nmultivariate time series. We extend the concept of Granger causality using\nprobabilistic forecasting with DeepAR. Since deep networks can neither handle\nmissing input nor out-of-distribution intervention, we propose to use the\nKnockoffs framework (Barberand Cand`es, 2015) for generating intervention\nvariables and consequently counterfactual probabilistic forecasting. Knockoff\nsamples are independent of their output given the observed variables and\nexchangeable with their counterpart variables without changing the underlying\ndistribution of the data. We test our method on synthetic as well as real-world\ntime series datasets. Overall our method outperforms the widely used vector\nautoregressive Granger causality and PCMCI in detecting nonlinear causal\ndependency in multivariate time series.\n", " Topological methods, including persistent homology, are powerful tools for\nanalysis of high-dimensional data sets but these methods rely almost\nexclusively on thresholding techniques. In noisy data sets, thresholding does\nnot always allow for the recovery of topological information. We present an\neasy to implement, computationally efficient pre-processing algorithm to\nprepare noisy point cloud data sets for topological data analysis. The\ntopological de-noising algorithm allows for the recovery of topological\ninformation that is inaccessible by thresholding methods. We apply the\nalgorithm to synthetically-generated noisy data sets and show the recovery of\ntopological information which is impossible to obtain by thresholding. We also\napply the algorithm to natural image data in R^8 and show a very clean recovery\nof topological information previously only available with large amounts of\nthresholding. Finally, we discuss future directions for improving this\nalgorithm using zig-zag persistence methods.\n", " In a previous paper [Pearl and Verma, 1991] we presented an algorithm for\nextracting causal influences from independence information, where a causal\ninfluence was defined as the existence of a directed arc in all minimal causal\nmodels consistent with the data. In this paper we address the question of\ndeciding whether there exists a causal model that explains ALL the observed\ndependencies and independencies. Formally, given a list M of conditional\nindependence statements, it is required to decide whether there exists a\ndirected acyclic graph (dag) D that is perfectly consistent with M, namely,\nevery statement in M, and no other, is reflected via dseparation in D. We\npresent and analyze an effective algorithm that tests for the existence of such\na day, and produces one, if it exists.\n", " Remarkable effectiveness of the channel or spatial attention mechanisms for\nproducing more discernible feature representation are illustrated in various\ncomputer vision tasks. However, modeling the cross-channel relationships with\nchannel dimensionality reduction may bring side effect in extracting deep\nvisual representations. In this paper, a novel efficient multi-scale attention\n(EMA) module is proposed. Focusing on retaining the information on per channel\nand decreasing the computational overhead, we reshape the partly channels into\nthe batch dimensions and group the channel dimensions into multiple\nsub-features which make the spatial semantic features well-distributed inside\neach feature group. Specifically, apart from encoding the global information to\nre-calibrate the channel-wise weight in each parallel branch, the output\nfeatures of the two parallel branches are further aggregated by a\ncross-dimension interaction for capturing pixel-level pairwise relationship. We\nconduct extensive ablation studies and experiments on image classification and\nobject detection tasks with popular benchmarks (e.g., CIFAR-100, ImageNet-1k,\nMS COCO and VisDrone2019) for evaluating its performance.\n", " This paper introduces Whittemore, a language for causal programming. Causal\nprogramming is based on the theory of structural causal models and consists of\ntwo primary operations: identification, which finds formulas that compute\ncausal queries, and estimation, which applies formulas to transform probability\ndistributions to other probability distribution. Causal programming provides\nabstractions to declare models, queries, and distributions with syntax similar\nto standard mathematical notation, and conducts rigorous causal inference,\nwithout requiring detailed knowledge of the underlying algorithms. Examples of\ncausal inference with real data are provided, along with discussion of the\nimplementation and possibilities for future extension.\n", " To mitigate errors induced by the cell's heterogeneous noisy environment, its\nmain information channels and production networks utilize the kinetic\nproofreading (KPR) mechanism. Here, we examine two extensively-studied KPR\ncircuits, DNA replication by the T7 DNA polymerase and translation by the E.\ncoli ribosome. Using experimental data, we analyze the performance of these two\nvital systems in light of the fundamental bounds set by the recently-discovered\nthermodynamic uncertainty relation (TUR), which places an inherent trade-off\nbetween the precision of a desirable output and the amount of energy\ndissipation required. We show that the DNA polymerase operates close to the TUR\nlower bound, while the ribosome operates $\\sim5$ times farther from this bound.\nThis difference originates from the enhanced binding discrimination of the\npolymerase which allows it to operate effectively as a reduced reaction cycle\nprioritizing correct product formation. We show that approaching this limit\nalso decouples the thermodynamic uncertainty factor from speed and error,\nthereby relaxing the accuracy-speed trade-off of the system. Altogether, our\nresults show that operating near this reduced cycle limit not only minimizes\nthermodynamic uncertainty, but also results in global performance enhancement\nof KPR circuits.\n", " The success of contextual word representations and advances in neural\ninformation retrieval have made dense vector-based retrieval a standard\napproach for passage and document ranking. While effective and efficient,\ndual-encoders are brittle to variations in query distributions and noisy\nqueries. Data augmentation can make models more robust but introduces overhead\nto training set generation and requires retraining and index regeneration. We\npresent Contrastive Alignment POst Training (CAPOT), a highly efficient\nfinetuning method that improves model robustness without requiring index\nregeneration, the training set optimization, or alteration. CAPOT enables\nrobust retrieval by freezing the document encoder while the query encoder\nlearns to align noisy queries with their unaltered root. We evaluate CAPOT\nnoisy variants of MSMARCO, Natural Questions, and Trivia QA passage retrieval,\nfinding CAPOT has a similar impact as data augmentation with none of its\noverhead.\n", " A new approach for the approximation of the channel log-likelihood ratio\n(LLR) for wireless channels based on Taylor series is proposed. The\napproximation is applied to the uncorrelated flat Rayleigh fading channel with\nunknown channel state information at the receiver. It is shown that the\nproposed approximation greatly simplifies the calculation of channel LLRs, and\nyet provides results almost identical to those based on the exact calculation\nof channel LLRs. The results are obtained in the context of bit-interleaved\ncoded modulation (BICM) schemes with low-density parity-check (LDPC) codes, and\ninclude threshold calculations and error rate performance of finite-length\ncodes. Compared to the existing approximations, the proposed method is either\nsignificantly less complex, or considerably more accurate.\n", " We show that two almost degenerate poles near the $\\pi\\Delta$ threshold and\nthe next higher mass pole in the $P_{11}$ partial wave of $\\pi N$ scattering\nevolve from a single bare state through its coupling with $\\pi N$, $\\eta N$ and\n$\\pi\\pi N$ reaction channels. This finding provides new information on\nunderstanding the dynamical origins of the Roper $N^*(1440)$ and $N^*(1710)$\nresonances listed by Particle Data Group. Our results for the resonance poles\nin other $\\pi N$ partial waves are also presented.\n", " In many scientific contexts, different investigators experiment with or\nobserve different variables with data from a domain in which the distinct\nvariable sets might well be related. This sort of fragmentation sometimes\noccurs in molecular biology, whether in studies of RNA expression or studies of\nprotein interaction, and it is common in the social sciences. Models are built\non the diverse data sets, but combining them can provide a more unified account\nof the causal processes in the domain. On the other hand, this problem is made\nchallenging by the fact that a variable in one data set may influence variables\nin another although neither data set contains all of the variables involved.\nSeveral authors have proposed using conditional independence properties of\nfragmentary (marginal) data collections to form unified causal explanations\nwhen it is assumed that the data have a common causal explanation but cannot be\nmerged to form a unified dataset. These methods typically return a large number\nof alternative causal models. The first part of the thesis shows that marginal\ndatasets contain extra information that can be used to reduce the number of\npossible models, in some cases yielding a unique model.\n", " Discovering causal effects is at the core of scientific investigation but\nremains challenging when only observational data is available. In practice,\ncausal networks are difficult to learn and interpret, and limited to relatively\nsmall datasets. We report a more reliable and scalable causal discovery method\n(iMIIC), based on a general mutual information supremum principle, which\ngreatly improves the precision of inferred causal relations while\ndistinguishing genuine causes from putative and latent causal effects. We\nshowcase iMIIC on synthetic and real-life healthcare data from 396,179 breast\ncancer patients from the US Surveillance, Epidemiology, and End Results\nprogram. More than 90\\% of predicted causal effects appear correct, while the\nremaining unexpected direct and indirect causal effects can be interpreted in\nterms of diagnostic procedures, therapeutic timing, patient preference or\nsocio-economic disparity. iMIIC's unique capabilities open up new avenues to\ndiscover reliable and interpretable causal networks across a range of research\nfields.\n", " In this paper, we introduce the notion of Plausible Deniability in an\ninformation theoretic framework. We consider a scenario where an entity that\neavesdrops through a broadcast channel summons one of the parties in a\ncommunication protocol to reveal their message (or signal vector). It is\ndesirable that the summoned party have enough freedom to produce a fake output\nthat is likely plausible given the eavesdropper's observation. We examine three\nvariants of this problem -- Message Deniability, Transmitter Deniability, and\nReceiver Deniability. In the first setting, the message sender is summoned to\nproduce the sent message. Similarly, in the second and third settings, the\ntransmitter and the receiver are required to produce the transmitted codeword,\nand the received vector respectively. For each of these settings, we examine\nthe maximum communication rate that allows a given minimum rate of plausible\nfake outputs. For the Message and Transmitter Deniability problems, we fully\ncharacterise the capacity region for general broadcast channels, while for the\nReceiver Deniability problem, we give an achievable rate region for physically\ndegraded broadcast channels.\n", " Partially-observed Boolean dynamical systems (POBDS) are a general class of\nnonlinear models with application in estimation and control of Boolean\nprocesses based on noisy and incomplete measurements. The optimal minimum mean\nsquare error (MMSE) algorithms for POBDS state estimation, namely, the Boolean\nKalman filter (BKF) and Boolean Kalman smoother (BKS), are intractable in the\ncase of large systems, due to computational and memory requirements. To address\nthis, we propose approximate MMSE filtering and smoothing algorithms based on\nthe auxiliary particle filter (APF) method from sequential Monte-Carlo theory.\nThese algorithms are used jointly with maximum-likelihood (ML) methods for\nsimultaneous state and parameter estimation in POBDS models. In the presence of\ncontinuous parameters, ML estimation is performed using the\nexpectation-maximization (EM) algorithm; we develop for this purpose a special\nsmoother which reduces the computational complexity of the EM algorithm. The\nresulting particle-based adaptive filter is applied to a POBDS model of Boolean\ngene regulatory networks observed through noisy RNA-Seq time series data, and\nperformance is assessed through a series of numerical experiments using the\nwell-known cell cycle gene regulatory model.\n", " Quantum state transfer plays an important role in quantum information\nprocessing. The evolution of certain pair states in a quantum network with\nHeisenberg $XY$ Hamiltonian depends only on the local structure of the network,\nand it remains unchanged even if the global structure is altered. All graphs\nwith high fidelity vertex state transfer may be considered as isomorphic\nbranches of the graph underlying a large quantum network to exhibit high\nfidelity pair state transfer. Among other graphs, one may construct infinite\nfamily of trees admitting perfect pair state transfer.\n", " Recent interest in integrated sensing and communications has led to the\ndesign of novel signal processing techniques to recover information from an\noverlaid radar-communications signal. Here, we focus on a spectral coexistence\nscenario, wherein the channels and transmit signals of both radar and\ncommunications systems are unknown to the common receiver. In this dual-blind\ndeconvolution (DBD) problem, the receiver admits a multi-carrier wireless\ncommunications signal that is overlaid with the radar signal reflected off\nmultiple targets. The communications and radar channels are represented by\ncontinuous-valued range-times or delays corresponding to multiple transmission\npaths and targets, respectively. Prior works addressed recovery of unknown\nchannels and signals in this ill-posed DBD problem through atomic norm\nminimization but contingent on individual minimum separation conditions for\nradar and communications channels. In this paper, we provide an optimal joint\nseparation condition using extremal functions from the Beurling-Selberg\ninterpolation theory. Thereafter, we formulate DBD as a low-rank modified\nHankel matrix retrieval and solve it via nuclear norm minimization. We estimate\nthe unknown target and communications parameters from the recovered low-rank\nmatrix using multiple signal classification (MUSIC) method. We show that the\njoint separation condition also guarantees that the underlying Vandermonde\nmatrix for MUSIC is well-conditioned. Numerical experiments validate our\ntheoretical findings.\n", " The availability of accurate localization is critical for multi-robot\nexploration strategies; noisy or inconsistent localization causes failure in\nmeeting exploration objectives. We aim to achieve high localization accuracy\nwith contemporary exploration map belief and vice versa without needing global\nlocalization information. This paper proposes a novel simultaneous exploration\nand localization (SEAL) approach, which uses Gaussian Processes (GP)-based\ninformation fusion for maximum exploration while performing communication graph\noptimization for relative localization. Both these cross-dependent objectives\nwere integrated through the Rao-Blackwellization technique. Distributed\nlinearized convex hull optimization is used to select the next-best unexplored\nregion for distributed exploration. SEAL outperformed cutting-edge methods on\nexploration and localization performance in extensive ROS-Gazebo simulations,\nillustrating the practicality of the approach in real-world applications.\n", " A proper parametrization of state transition matrices of linear state-space\nmodels (SSMs) followed by standard nonlinearities enables them to efficiently\nlearn representations from sequential data, establishing the state-of-the-art\non a large series of long-range sequence modeling benchmarks. In this paper, we\nshow that we can improve further when the structural SSM such as S4 is given by\na linear liquid time-constant (LTC) state-space model. LTC neural networks are\ncausal continuous-time neural networks with an input-dependent state transition\nmodule, which makes them learn to adapt to incoming inputs at inference. We\nshow that by using a diagonal plus low-rank decomposition of the state\ntransition matrix introduced in S4, and a few simplifications, the LTC-based\nstructural state-space model, dubbed Liquid-S4, achieves the new\nstate-of-the-art generalization across sequence modeling tasks with long-term\ndependencies such as image, text, audio, and medical time-series, with an\naverage performance of 87.32% on the Long-Range Arena benchmark. On the full\nraw Speech Command recognition, dataset Liquid-S4 achieves 96.78% accuracy with\na 30% reduction in parameter counts compared to S4. The additional gain in\nperformance is the direct result of the Liquid-S4's kernel structure that takes\ninto account the similarities of the input sequence samples during training and\ninference.\n", " We present a definition of cause and effect in terms of decision-theoretic\nprimitives and thereby provide a principled foundation for causal reasoning.\nOur definition departs from the traditional view of causation in that causal\nassertions may vary with the set of decisions available. We argue that this\napproach provides added clarity to the notion of cause. Also in this paper, we\nexamine the encoding of causal relationships in directed acyclic graphs. We\ndescribe a special class of influence diagrams, those in canonical form, and\nshow its relationship to Pearl's representation of cause and effect. Finally,\nwe show how canonical form facilitates counterfactual reasoning.\n", " Causal gene networks model the flow of information within a cell, but\nreconstructing them from omics data is challenging because correlation does not\nimply causation. Combining genomics and transcriptomics data from a segregating\npopulation allows to orient the direction of causality between gene expression\ntraits using genomic variants. Instrumental-variable methods (IV) use a local\nexpression quantitative trait locus (eQTL) as a randomized instrument for a\ngene's expression level, and assign target genes based on distal eQTL\nassociations. Mediation-based methods (ME) additionally require that distal\neQTL associations are mediated by the source gene. Here we used Findr, a\nsoftware providing uniform implementations of IV, ME, and coexpression-based\nmethods, a recent dataset of 1,012 segregants from a cross between two budding\nyeast strains, and the YEASTRACT database of known transcriptional interactions\nto compare causal gene network inference methods. We found that causal\ninference methods result in a significant overlap with the ground-truth,\nwhereas coexpression did not perform better than random. A subsampling analysis\nrevealed that the performance of ME decreases at large sample sizes, due to a\nloss of sensitivity when residual correlations become significant. IV methods\ncontain false positive predictions, due to genomic linkage between eQTL\ninstruments. IV and ME methods also have complementary roles for identifying\ncausal genes underlying transcriptional hotspots. IV methods correctly\npredicted STB5 targets for a hotspot centred on the transcription factor STB5,\nwhereas ME failed due to Stb5p auto-regulating its own expression. ME suggests\na new candidate gene, DNM1, for a hotspot on Chr XII, where IV methods could\nnot distinguish between multiple genes located within the hotspot.\n", " To clarify some aspects of the application of Special Relativity, spacetime\nis sliced into null geodesic hypersurfaces as an alternative to the\nhypersurfaces of simultaneity normally adopted. Events at particle locations on\nthe hypersurface are identified as the causal event set. It is demonstrated\nthat a Lorentz boost applied to the causal event set maintains the property of\nconnectedness and with this formalism it is simple to derive the redshift\nequation. The twin paradox is naturally explained as an instantaneous\nreconfiguration of particle position 4-vectors in the frame of the accelerated\nobject. The metaphysical implications are examined with the tentative\nconclusion that a relationist view of spacetime is more consistent with this\ntreatment than the substantivalist viewpoint.\n", " In this work we study the first step in photosynthesis for the limiting case\nof a single photon interacting with photosystem II (PSII). We model our system\nusing quantum trajectory theory, which allows us to consider not only the\naverage evolution, but also the conditional evolution of the system given\nindividual realizations of idealized measurements of photons that have been\nabsorbed and subsequently emitted as fluorescence. The quantum nature of the\nsingle photon input requires a fully quantum model of both the input and output\nlight fields. We show that PSII coupled to the field via three collective\n``bright states'', whose orientation and distribution correlate strongly with\nits natural geometry. Measurements of the transmitted beam strongly affects the\nsystem state, since a (null) detection of the outgoing photon confirms that the\nsystem must be in the electronic (excited) ground state. Using numerical and\nanalytical calculations we show that observing the null result transforms a\nstate with a low excited state population $O( 10^{-5} )$ to a state with nearly\nall population contained in the excited states. This is solely a property of\nthe single photon input, as we confirm by comparing this behavior with that for\nexcitation by a coherent state possessing an average of one photon, using a\nsmaller five site ``pentamer'' system. We also examine the effect of a\ndissipative phononic environment on the conditional excited state dynamics. We\nshow that the environment has a strong effect on the observed rates of\nfluorescence, which could act as a new photon-counting witness of excitonic\ncoherence. The long time evolution of the phononic model predicts an\nexperimentally consistent quantum efficiency of 92%.\n", " Understanding relationship between noisy dynamics and biological network\narchitecture is a fundamentally important question, particularly in order to\nelucidate how cells encode and process information. We analytically and\nnumerically investigate general network architectural conditions that are\nnecessary to generate stochastic amplified and coherent oscillations. We\nenumerate all possible topologies of coupled negative feedbacks in the\nunderlying biochemical networks with three components, negative feedback loops,\nand mass action kinetics. Using the linear noise approximation to analytically\nobtain the time-dependent solution of the master equation and derive the\nalgebraic expression of power spectra, we find that (a) all networks with\ncoupled negative feedbacks are capable of generating stochastic amplified and\ncoherent oscillations; (b) networks with a single negative feedback are better\nstochastic amplified and coherent oscillators than those with multiple coupled\nnegative feedbacks; (c) multiple timescale difference among the kinetic rate\nconstants is required for stochastic amplified and coherent oscillations.\n", " Probabilistic Boolean Networks have been proposed for estimating the\nbehaviour of dynamical systems as they combine rule-based modelling with\nuncertainty principles. Inferring PBNs directly from gene data is challenging\nhowever, especially when data is costly to collect and/or noisy, e.g., in the\ncase of gene expression profile data. In this paper, we present a reproducible\nmethod for inferring PBNs directly from real gene expression data measurements\ntaken when the system was at a steady state. The steady-state dynamics of PBNs\nis of special interest in the analysis of biological machinery. The proposed\napproach does not rely on reconstructing the state evolution of the network,\nwhich is computationally intractable for larger networks. We demonstrate the\nmethod on samples of real gene expression profiling data from a well-known\nstudy on metastatic melanoma. The pipeline is implemented using Python and we\nmake it publicly available.\n", " Causal structure learning has been extensively studied and widely used in\nmachine learning and various applications. To achieve an ideal performance,\nexisting causal structure learning algorithms often need to centralize a large\namount of data from multiple data sources. However, in the privacy-preserving\nsetting, it is impossible to centralize data from all sources and put them\ntogether as a single dataset. To preserve data privacy, federated learning as a\nnew learning paradigm has attracted much attention in machine learning in\nrecent years. In this paper, we study a privacy-aware causal structure learning\nproblem in the federated setting and propose a novel Federated PC (FedPC)\nalgorithm with two new strategies for preserving data privacy without\ncentralizing data. Specifically, we first propose a novel layer-wise\naggregation strategy for a seamless adaptation of the PC algorithm into the\nfederated learning paradigm for federated skeleton learning, then we design an\neffective strategy for learning consistent separation sets for federated edge\norientation. The extensive experiments validate that FedPC is effective for\ncausal structure learning in a federated learning setting.\n", " This paper presents a technique to train a robot to perform kick-motion in AI\nsoccer by using reinforcement learning (RL). In RL, an agent interacts with an\nenvironment and learns to choose an action in a state at each step. When\ntraining RL algorithms, a problem called the curse of dimensionality (COD) can\noccur if the dimension of the state is high and the number of training data is\nlow. The COD often causes degraded performance of RL models. In the situation\nof the robot kicking the ball, as the ball approaches the robot, the robot\nchooses the action based on the information obtained from the soccer field. In\norder not to suffer COD, the training data, which are experiences in the case\nof RL, should be collected evenly from all areas of the soccer field over\n(theoretically infinite) time. In this paper, we attempt to use the relative\ncoordinate system (RCS) as the state for training kick-motion of robot agent,\ninstead of using the absolute coordinate system (ACS). Using the RCS eliminates\nthe necessity for the agent to know all the (state) information of entire\nsoccer field and reduces the dimension of the state that the agent needs to\nknow to perform kick-motion, and consequently alleviates COD. The training\nbased on the RCS is performed with the widely used Deep Q-network (DQN) and\ntested in the AI Soccer environment implemented with Webots simulation\nsoftware.\n", " Despite large-scale pre-trained language models have achieved striking\nresults for text classificaion, recent work has raised concerns about the\nchallenge of shortcut learning. In general, a keyword is regarded as a shortcut\nif it creates a superficial association with the label, resulting in a false\nprediction. Conversely, shortcut learning can be mitigated if the model relies\non robust causal features that help produce sound predictions. To this end,\nmany studies have explored post-hoc interpretable methods to mine shortcuts and\ncausal features for robustness and generalization. However, most existing\nmethods focus only on single word in a sentence and lack consideration of\nword-group, leading to wrong causal features. To solve this problem, we propose\na new Word-Group mining approach, which captures the causal effect of any\nkeyword combination and orders the combinations that most affect the\nprediction. Our approach bases on effective post-hoc analysis and beam search,\nwhich ensures the mining effect and reduces the complexity. Then, we build a\ncounterfactual augmentation method based on the multiple word-groups, and use\nan adaptive voting mechanism to learn the influence of different augmentated\nsamples on the prediction results, so as to force the model to pay attention to\neffective causal features. We demonstrate the effectiveness of the proposed\nmethod by several tasks on 8 affective review datasets and 4 toxic language\ndatasets, including cross-domain text classificaion, text attack and gender\nfairness test.\n", " Many applications, including rank aggregation, crowd-labeling, and graphon\nestimation, can be modeled in terms of a bivariate isotonic matrix with unknown\npermutations acting on its rows and/or columns. We consider the problem of\nestimating an unknown matrix in this class, based on noisy observations of\n(possibly, a subset of) its entries. We design and analyze polynomial-time\nalgorithms that improve upon the state of the art in two distinct metrics,\nshowing, in particular, that minimax optimal, computationally efficient\nestimation is achievable in certain settings. Along the way, we prove matching\nupper and lower bounds on the minimax radii of certain cone testing problems,\nwhich may be of independent interest.\n", " Electron transport in realistic physical and chemical systems often involves\nthe non-trivial exchange of energy with a large environment, requiring the\ndefinition and treatment of open quantum systems. Because the time evolution of\nan open quantum system employs a non-unitary operator, the simulation of open\nquantum systems presents a challenge for universal quantum computers\nconstructed from only unitary operators or gates. Here we present a general\nalgorithm for implementing the action of any non-unitary operator on an\narbitrary state on a quantum device. We show that any quantum operator can be\nexactly decomposed as a linear combination of at most four unitary operators.\nWe demonstrate this method on a two-level system in both zero and finite\ntemperature amplitude damping channels. The results are in agreement with\nclassical calculations, showing promise in simulating non-unitary operations on\nintermediate-term and future quantum devices.\n", " Reinforcement learning (RL) is a promising approach to solve dialogue policy\noptimisation. Traditional RL algorithms, however, fail to scale to large\ndomains due to the curse of dimensionality. We propose a novel Dialogue\nManagement architecture, based on Feudal RL, which decomposes the decision into\ntwo steps; a first step where a master policy selects a subset of primitive\nactions, and a second step where a primitive action is chosen from the selected\nsubset. The structural information included in the domain ontology is used to\nabstract the dialogue state space, taking the decisions at each step using\ndifferent parts of the abstracted state. This, combined with an information\nsharing mechanism between slots, increases the scalability to large domains. We\nshow that an implementation of this approach, based on Deep-Q Networks,\nsignificantly outperforms previous state of the art in several dialogue domains\nand environments, without the need of any additional reward signal.\n", " Perfect state transfer and fractional revival can be used to move information\nbetween pairs of vertices in a quantum network. While perfect state transfer\nhas received a lot of attention, fractional revival is newer and less studied.\nOne problem is to determine the differences between perfect state transfer and\nfractional revival. If perfect state transfer occurs between two vertices in a\ngraph, the vertices must be cospectral. Further if there is perfect state\ntransfer between vertices $a$ and $b$ in a graph, there cannot be perfect state\ntransfer from $a$ to any other vertex. No examples of unweighted graphs with\nfractional revival between non-cospectral vertices were known; here we give an\ninfinite family of such graphs. No examples of unweighted graphs where the\npairs involved in fractional revival overlapped were known; we give examples of\nsuch graphs as well.\n", " Analog beamforming is an attractive and cost-effective solution to exploit\nthe benefits of massive multiple-input-multiple-output systems, by requiring\nonly one up/down-conversion chain. However, the presence of only one chain\nimposes a significant overhead in estimating the channel state information\nrequired for beamforming, when conventional digital channel estimation (CE)\napproaches are used. As an alternative, this paper proposes a novel CE\ntechnique, called periodic analog CE (PACE), that can be performed by analog\nhardware. By avoiding digital processing, the estimation overhead is\nsignificantly lowered and does not scale with number of antennas. PACE involves\nperiodic transmission of a sinusoidal reference signal by the transmitter,\nestimation of its amplitude and phase at each receive antenna via analog\nhardware, and using these estimates for beamforming. To enable such non-trivial\noperation, two reference tone recovery techniques and a novel receiver\narchitecture for PACE are proposed and analyzed, both theoretically and via\nsimulations. Results suggest that in sparse, wide-band channels and above a\ncertain signal-to-noise ratio, PACE aided beamforming suffers only a small loss\nin beamforming gain and enjoys a much lower CE overhead, in comparison to\nconventional approaches. Benefits of using PACE aided beamforming during the\ninitial access phase are also discussed.\n", " The magnetotransport properties of massless Dirac fermions in a gapless HgTe\nquantum well are investigated. In samples with narrow channels, a large\nnegative magnetoresistance with a Lorentzian profile is observed, which is\ninterpreted as a manifestation of electron viscosity due to electron-electron\ninteraction. Comparison of experiment with theory yields the shear stress\nrelaxation time of the Dirac fermions caused by electron-electron scattering.\n", " Gene regulatory network inference (GRNI) is a challenging problem,\nparticularly owing to the presence of zeros in single-cell RNA sequencing data:\nsome are biological zeros representing no gene expression, while some others\nare technical zeros arising from the sequencing procedure (aka dropouts), which\nmay bias GRNI by distorting the joint distribution of the measured gene\nexpressions. Existing approaches typically handle dropout error via imputation,\nwhich may introduce spurious relations as the true joint distribution is\ngenerally unidentifiable. To tackle this issue, we introduce a causal graphical\nmodel to characterize the dropout mechanism, namely, Causal Dropout Model. We\nprovide a simple yet effective theoretical result: interestingly, the\nconditional independence (CI) relations in the data with dropouts, after\ndeleting the samples with zero values (regardless if technical or not) for the\nconditioned variables, are asymptotically identical to the CI relations in the\noriginal data without dropouts. This particular test-wise deletion procedure,\nin which we perform CI tests on the samples without zeros for the conditioned\nvariables, can be seamlessly integrated with existing structure learning\napproaches including constraint-based and greedy score-based methods, thus\ngiving rise to a principled framework for GRNI in the presence of dropouts. We\nfurther show that the causal dropout model can be validated from data, and many\nexisting statistical models to handle dropouts fit into our model as specific\nparametric instances. Empirical evaluation on synthetic, curated, and\nreal-world experimental transcriptomic data comprehensively demonstrate the\nefficacy of our method.\n", " In this paper, we study a long-term planning scenario that is based on drone\nracing competitions held in real life. We conducted this experiment on a\nframework created for \"Game of Drones: Drone Racing Competition\" at NeurIPS\n2019. The racing environment was created using Microsoft's AirSim Drone Racing\nLab. A reinforcement learning agent, a simulated quadrotor in our case, has\ntrained with the Policy Proximal Optimization(PPO) algorithm was able to\nsuccessfully compete against another simulated quadrotor that was running a\nclassical path planning algorithm. Agent observations consist of data from IMU\nsensors, GPS coordinates of drone obtained through simulation and opponent\ndrone GPS information. Using opponent drone GPS information during training\nhelps dealing with complex state spaces, serving as expert guidance allows for\nefficient and stable training process. All experiments performed in this paper\ncan be found and reproduced with code at our GitHub repository\n", " Geometric graphs form an important family of hidden structures behind data.\nIn this paper, we develop an efficient and robust algorithm to infer a graph\nskeleton of a high-dimensional point cloud dataset (PCD). Previously, there has\nbeen much work to recover a hidden graph from a low-dimensional density field,\nor from a relatively clean high-dimensional PCD. Our proposed approach builds\nupon the recent line of work on using a persistence-guided discrete Morse (DM)\ntheory based approach to reconstruct a geometric graph from a density field\ndefined over a low-dimensional triangulation. In particular, we first give a\nvery simple generalization of this DM-based algorithm from a density-function\nperspective to a general filtration perspective. On the theoretical front, we\nshow that the output of the generalized algorithm contains a so-called\nlexicographic-optimal persistent cycle basis w.r.t the input filtration,\njustifying that the output is indeed meaningful. On the algorithmic front, the\ngeneralization allows us to combine sparsified weighted Rips filtration to\ndevelop a new graph reconstruction algorithm for noisy point cloud data. The\nnew algorithm is robust to background noise and non-uniform distribution of\ninput points, and we provide various experimental results to show its\neffectiveness.\n", " Polymer Optical Fiber (POF) can be and are being used in various fields of\napplications. Two of the main fields are the automotive and the home\nentertainment sector. The POF can be applied in several different optical\ncommunication systems as automotive multi-media busses or in-house Ethernet\nsystems.\n The requirements of bandwidth are increasing very fast in these sectors and\ntherefore solutions that satisfy these demands are of high actuality. One\nsolution is to use the wavelength division multiplexing (WDM) technique. Here,\nseveral different wavelengths can carry information over one POF fiber. All\nwavelengths that are transmitted over the fiber, must be separated at the\nreceiver to regain and redirect the information channels. These separators are\nso-called Demultiplexers.\n There are several systems available on the market, which are all afflicted\nwith certain disadvantages. But all these solutions have one main disadvantage,\nthey are all too expensive for most of the applications mentioned above. So the\ngoal of this study is to develop an economical Demultiplexer for WDM\ntransmission over POF.\n The main idea is to separate the chromatic light in its monochromatic\ncomponents with the help of a prism with low reciprocal dispersive power. The\nprism and the other assemblies, which are needed to adjust the optical path,\nshould be manufactured in injection molding technique. This manufacturing\ntechnique is a very simple and economical way to produce a mass production\napplicable Demultiplexer for POF.\n", " Data Science aims to extract meaningful knowledge from unorganised data. Real\ndatasets usually come in the form of a cloud of points with only pairwise\ndistances. Numerous applications require to visualise an overall shape of a\nnoisy cloud of points sampled from a non-linear object that is more complicated\nthan a union of disjoint clusters. The skeletonisation problem in its hardest\nform is to find a 1-dimensional skeleton that correctly represents a shape of\nthe cloud. This paper compares several algorithms that solve the above\nskeletonisation problem for any point cloud and guarantee a successful\nreconstruction. For example, given a highly noisy point sample of an unknown\nunderlying graph, a reconstructed skeleton should be geometrically close and\nhomotopy equivalent to (has the same number of independent cycles as) the\nunderlying graph. One of these algorithm produces a Homologically Persistent\nSkeleton (HoPeS) for any cloud without extra parameters. This universal\nskeleton contains sub-graphs that provably represent the 1-dimensional shape of\nthe cloud at any scale. Other subgraphs of HoPeS reconstruct an unknown graph\nfrom its noisy point sample with a correct homotopy type and within a small\noffset of the sample. The extensive experiments on synthetic and real data\nreveal for the first time the maximum level of noise that allows successful\ngraph reconstructions.\n", " Computing based on biochemical processes is a newest rapidly developing field\nof unconventional information and signal processing. In this paper we present\nresults of our research in the field of biochemical computing and summarize the\nobtained numerical and experimental data for implementations of the standard\ntwo-input OR and AND gates with double-sigmoid shape of the output signal. This\nform of response was obtained as a function of the two inputs in each of the\nrealized biochemical systems. The enzymatic gate processes in the first system\nwere activated with two chemical inputs and resulted in optically detected\nchromogen oxidation, which happens when either one or both of the inputs are\npresent. In this case, the biochemical system is functioning as the OR gate. We\ndemonstrate that the addition of a \"filtering\" biocatalytic process leads to a\nconsiderable reduction of the noise transmission factor and the resulting gate\nresponse has sigmoid shape in both inputs. The second system was developed for\nfunctioning as an AND gate, where the output signal was activated only by a\nsimultaneous action of two enzymatic biomarkers. This response can be used as\nan indicator of liver damage, but only if both of these of the inputs are\npresent at their elevated, pathophysiological values of concentrations. A\nkinetic numerical model was developed and used to estimate the range of\nparameters for which the experimentally realized logic gate is close to\noptimal. We also analyzed the system to evaluate its noise-handling properties.\n", " We present a model for neutrino-nucleus scattering in the energy region\nrelevant for present and forthcoming neutrino-oscillation experiments. The\nmodel is based on the RPA treatment of the nuclear responses in the\nquasi-elastic and Delta-resonance region. It includes also in a\nphenomenological way nucleon knock-out. It aims at the description, within a\nsingle framework, of several final state channels i.e. quasi-elastic,\nincoherent and coherent one-pion production and two- or several-nucleon\nknock-out.\n", " The Fortran code NMHDECAY computes the masses, couplings and decay widths of\nall Higgs bosons of the NMSSM in terms of its parameters at the electroweak\n(SUSY breaking) scale: the Yukawa couplings lambda and kappa, the soft\ntrilinear terms A_lambda and A_kappa, and tan(beta) and mu_eff = lambda*<S>.\nThe computation of the spectrum includes leading two loop terms, electroweak\ncorrections and propagator corrections. The computation of the decay widths is\ncarried out as in HDECAY, but (for the moment) without three body decays. Each\npoint in parameter space is checked against negative Higgs bosons searches at\nLEP, including unconventional channels relevant for the NMSSM. One version of\nthe program uses generalized SLHA conventions for input and output.\n", " In this paper we show how inter-cellular molecular communication may change\nthe overall levels of photosynthesis in plants. Individual plant cells respond\nto external stimuli, such as illumination levels, to regulate their\nphotosynthetic output. Here, we present a mathematical model which shows that\nby sharing information internally using molecular communication, plants may\nincrease overall photosynthate production. Numerical results show that higher\nmutual information between cells corresponds to an increase in overall\nphotosynthesis by as much as 25 per cent. This suggests that molecular\ncommunication plays a vital role in maximising the photosynthesis in plants and\ntherefore suggests new routes to influence plant development in agriculture and\nelsewhere.\n", " Circadian clocks are biochemical oscillators that allow organisms to estimate\nthe time of the day. These oscillators are inherently noisy due to the discrete\nnature of the reactants and the stochastic character of their interactions. To\nkeep these oscillators in sync with the daily day-night rhythm in the presence\nof noise, circadian clocks must be coupled to the dark-light cycle. In this\nmanuscript, we study the entrainment of phase oscillators as a function of the\nintrinsic noise in the system. Using stochastic simulations, we compute the\noptimal coupling strength, intrinsic frequency and shape of the phase-response\ncurve, that maximize the mutual information between the phase of the clock and\ntime. We show that the optimal coupling strength and intrinsic frequency\nincrease with the noise, but that the shape of the phase-response curve varies\nnon-monotonically with the noise: in the low-noise regime, it features a\ndeadzone that increases in width as the noise increases, while in the\nhigh-noise regime, the width decreases with the noise. These results arise from\na trade-off between maximizing stability -noise suppression - and maximizing\nlinearity of the input-output, i.e. time- phase, relation. We also show that\nthree analytic approximations -the linear-noise approximation, the\nphase-averaging method, and linear-response theory- accurately describe\ndifferent regimes of the coupling strength and the noise.\n", " Detecting commonsense causal relations (causation) between events has long\nbeen an essential yet challenging task. Given that events are complicated, an\nevent may have different causes under various contexts. Thus, exploiting\ncontext plays an essential role in detecting causal relations. Meanwhile,\nprevious works about commonsense causation only consider two events and ignore\ntheir context, simplifying the task formulation. This paper proposes a new task\nto detect commonsense causation between two events in an event sequence (i.e.,\ncontext), called contextualized commonsense causal reasoning. We also design a\nzero-shot framework: COLA (Contextualized Commonsense Causality Reasoner) to\nsolve the task from the causal inference perspective. This framework obtains\nrich incidental supervision from temporality and balances covariates from\nmultiple timestamps to remove confounding effects. Our extensive experiments\nshow that COLA can detect commonsense causality more accurately than baselines.\n", " Trapped atoms and ions are among the best controlled quantum systems which\nfind widespread applications in quantum information, sensing and metrology. For\nmolecules, however, a similar degree of control is currently lacking owing to\ntheir complex energy-level structure. Quantum-logic protocols in which atomic\nions serve as probes for molecular ions are a promising route for achieving\nthis level of control, especially with homonuclear molecules that decouple from\nblack-body radiation. Here, a quantum-non-demolition protocol on single trapped\nN$_2^+$ molecules is demonstrated. The spin-rovibronic state of the molecule is\ndetected with more than 99% fidelity and the position and strength of a\nspectroscopic transition in the molecule are determined, both without\ndestroying the molecular quantum state. The present method lays the foundations\nfor new approaches to molecular precision spectroscopy, for state-to-state\nchemistry on the single-molecule level and for the implementation of molecular\nqubits.\n", " We consider two specific thermodynamic cycles of engine operating in a finite\ntime coupled to two thermal reservoirs with a finite heat capacity: The\nCarnot-type cycle and the Lorenz-type cycle. By means of the endo-reversible\nthermodynamics, we then discuss the power output of engine and its\noptimization. In doing so, we treat the temporal duration of a single cycle\nrigorously, i.e., without neglecting the duration of its adiabatic parts. Then\nwe find that the maximally obtainable power output P_m and the engine\nefficiency \\eta_m at the point of P_m explicitly depend on the heat conductance\nand the compression ratio. From this, it is immediate to observe that the\nwell-known results available in many references, in particular the\n(compression-ratio-independent) Curzon-Ahlborn-Novikov expressions such as\n\\eta_m --> \\eta_{CAN} = 1 - (T_L/T_H)^(1/2) with the temperatures (T_H, T_L) of\nhot and cold reservoirs only, can be recovered, but, significantly enough, in\nthe limit of a vanishingly small heat conductance and an infinitely large\ncompression ratio only. Our result implies that the endo-reversible model of a\nthermal machine operating in a finite time and so producing a finite power\noutput with the Curzon-Ahlborn-Novikov results should be limited in its own\nvalidity regime.\n", " In this paper, we present two software packages, HexGen and Hex2Spline, that\nseamlessly integrate geometry design with isogeometric analysis (IGA) in\nLS-DYNA. Given a boundary representation of a solid model, HexGen creates a\nhexahedral mesh by utilizing a semi-automatic polycube-based mesh generation\nmethod. Hex2Spline takes the output hexahedral mesh from HexGen as the input\ncontrol mesh and constructs volumetric truncated hierarchical splines. Through\nB\\'{e}zier extraction, Hex2Spline transfers spline information to LS-DYNA and\nperforms IGA therein. We explain the underlying algorithms in each software\npackage and use a rod model to explain how to run the software. We also apply\nour software to several other complex models to test its robustness. Our goal\nis to provide a robust volumetric modeling tool and thus expand the boundary of\nIGA to volume-based industrial applications.\n", " In this paper, we study a stochastic disclosure control problem using\ninformation-theoretic methods. The useful data to be disclosed depend on\nprivate data that should be protected. Thus, we design a privacy mechanism to\nproduce new data which maximizes the disclosed information about the useful\ndata under a strong $\\chi^2$-privacy criterion. For sufficiently small leakage,\nthe privacy mechanism design problem can be geometrically studied in the space\nof probability distributions by a local approximation of the mutual\ninformation. By using methods from Euclidean information geometry, the original\nhighly challenging optimization problem can be reduced to a problem of finding\nthe principal right-singular vector of a matrix, which characterizes the\noptimal privacy mechanism. In two extensions we first consider a scenario where\nan adversary receives a noisy version of the user's message and then we look\nfor a mechanism which finds $U$ based on observing $X$, maximizing the mutual\ninformation between $U$ and $Y$ while satisfying the privacy criterion on $U$\nand $Z$ under the Markov chain $(Z,Y)-X-U$.\n", " The intricacy of 3D surfaces often results cutting-edge point cloud denoising\n(PCD) models in surface degradation including remnant noise, wrongly-removed\ngeometric details. Although using multi-scale patches to encode the geometry of\na point has become the common wisdom in PCD, we find that simple aggregation of\nextracted multi-scale features can not adaptively utilize the appropriate scale\ninformation according to the geometric information around noisy points. It\nleads to surface degradation, especially for points close to edges and points\non complex curved surfaces. We raise an intriguing question -- if employing\nmulti-scale geometric perception information to guide the network to utilize\nmulti-scale information, can eliminate the severe surface degradation problem?\nTo answer it, we propose a Multi-offset Denoising Network (MODNet) customized\nfor multi-scale patches. First, we extract the low-level feature of three\nscales patches by patch feature encoders. Second, a multi-scale perception\nmodule is designed to embed multi-scale geometric information for each scale\nfeature and regress multi-scale weights to guide a multi-offset denoising\ndisplacement. Third, a multi-offset decoder regresses three scale offsets,\nwhich are guided by the multi-scale weights to predict the final displacement\nby weighting them adaptively. Experiments demonstrate that our method achieves\nnew state-of-the-art performance on both synthetic and real-scanned datasets.\n", " The fact that first-person conscious perceptions or sentient experiences have\nmany bits of information strongly suggests that they are produced nonlocally by\nthe effects of many atoms, say by nonlocal quantum operators. If these nonlocal\noperators act back on the quantum state of the atoms, they could lead to\nevolution in which signals propagate superluminally, violating the usual\ncausality of local quantum field theory. Although there is not yet any direct\nevidence that nonlocal operators associated with psycho-physical parallelism\nact back on the quantum state, it is not totally implausible that this might be\nthe case. In principle the resulting superluminal propagation might be\nobservable by sending signals across brain regions (neural correlates of\nconsciousness) that lead to conscious perceptions.\n" ]

[ " We give the first differentially private algorithms that estimate a variety\nof geometric features of points in the Euclidean space, such as diameter,\nwidth, volume of convex hull, min-bounding box, min-enclosing ball etc. Our\nwork relies heavily on the notion of \\emph{Tukey-depth}. Instead of\n(non-privately) approximating the convex-hull of the given set of points $P$,\nour algorithms approximate the geometric features of the $\\kappa$-Tukey region\ninduced by $P$ (all points of Tukey-depth $\\kappa$ or greater). Moreover, our\napproximations are all bi-criteria: for any geometric feature $\\mu$ our\n$(\\alpha,\\Delta)$-approximation is a value \"sandwiched\" between\n$(1-\\alpha)\\mu(D_P(\\kappa))$ and $(1+\\alpha)\\mu(D_P(\\kappa-\\Delta))$.\n Our work is aimed at producing a \\emph{$(\\alpha,\\Delta)$-kernel of\n$D_P(\\kappa)$}, namely a set $\\mathcal{S}$ such that (after a shift) it holds\nthat $(1-\\alpha)D_P(\\kappa)\\subset \\mathsf{CH}(\\mathcal{S}) \\subset\n(1+\\alpha)D_P(\\kappa-\\Delta)$. We show that an analogous notion of a bi-critera\napproximation of a directional kernel, as originally proposed by Agarwal et\nal~[2004], \\emph{fails} to give a kernel, and so we result to subtler notions\nof approximations of projections that do yield a kernel. First, we give\ndifferentially private algorithms that find $(\\alpha,\\Delta)$-kernels for a\n\"fat\" Tukey-region. Then, based on a private approximation of the min-bounding\nbox, we find a transformation that does turn $D_P(\\kappa)$ into a \"fat\" region\n\\emph{but only if} its volume is proportional to the volume of\n$D_P(\\kappa-\\Delta)$. Lastly, we give a novel private algorithm that finds a\ndepth parameter $\\kappa$ for which the volume of $D_P(\\kappa)$ is comparable to\n$D_P(\\kappa-\\Delta)$. We hope this work leads to the further study of the\nintersection of differential privacy and computational geometry.\n", " Changes in travel modes used by people, particularly reduction of the private\ncar use, is an important determinant of effectiveness of transportation plans.\nBecause of dependencies between the choices of residential location and travel\nmode, integrated modelling of these choices has been proposed by some\nresearchers. In this paper, an agent-based microsimulation model has been\ndeveloped to evaluate impacts of different transport development plans on\nchoices of residential location and commuting mode of tenant households in\nTehran, the capital of Iran. In the proposed model, households are considered\nas agents who select their desired residential location using a constrained\nNSGA-II algorithm and in a competition with other households. In addition, they\nchoose their commuting mode by applying a multi-criteria decision making\nmethod. Afterwards, effects of development of a new highway, subway and bus\nrapid transit (BRT) line on their residential location and commuting mode\nchoices are evaluated. Results show that despite the residential self-selection\neffects, these plans result in considerable changes in the commuting mode of\ndifferent socioeconomic categories of households. Development of the new subway\nline shows promising results by reducing the private car use among the all\nsocio-economic categories of households. But the new highway development\nunsatisfactorily results in increase in the private car use. In addition,\ndevelopment of the new BRT line does not show significant effects on the\ncommuting mode change, particularly on decrease in the private car use.\n", " Irredundance has been studied in the context of dominating sets, via the\nconcept of private neighbor. Here irredundance of zero forcing sets is\nintroduced via the concept of a private fort and the upper and lower zero\nforcing irrdedundance numbers $\\mbox{ZIR}(G)$ and $\\mbox{zir}(G)$ are defined.\nBounds on $\\mbox{ZIR}(G)$ and $\\mbox{zir}(G)$ are established and graphs having\nextreme values of $\\mbox{ZIR}(G)$ and $\\mbox{zir}(G)$ are characterized. The\neffect of the join and corona operations is studied. As the concept of a zero\nforcing irrdedundant set is new, there are many questions for future research.\n", " We study fractional smoothness of measures on $\\mathbb{R}^k$, that are images\nof a Gaussian measure under mappings from Gaussian Sobolev classes. As a\nconsequence we obtain Nikolskii--Besov fractional regularity of these\ndistributions under some weak nondegeneracy assumption.\n", " In this note we study the number of real roots of a wide class of random\northogonal polynomials with gaussian coefficients. Using the method of Wiener\nChaos we show that the fluctuation in the bulk is asymptotically gaussian, even\nwhen the local correlations are not necessarily the same.\n", " Consider the following distributed optimization scenario. A worker has access\nto training data that it uses to compute the gradients while a server decides\nwhen to stop iterative computation based on its target accuracy or delay\nconstraints. The server receives all its information about the problem instance\nfrom the worker via a rate-limited noiseless communication channel. We\nintroduce the principle we call Differential Quantization (DQ) that prescribes\ncompensating the past quantization errors to direct the descent trajectory of a\nquantized algorithm towards that of its unquantized counterpart. Assuming that\nthe objective function is smooth and strongly convex, we prove that\nDifferentially Quantized Gradient Descent (DQ-GD) attains a linear contraction\nfactor of $\\max\\{\\sigma_{\\mathrm{GD}}, \\rho_n 2^{-R}\\}$, where\n$\\sigma_{\\mathrm{GD}}$ is the contraction factor of unquantized gradient\ndescent (GD), $\\rho_n \\geq 1$ is the covering efficiency of the quantizer, and\n$R$ is the bitrate per problem dimension $n$. Thus at any $R\\geq\\log_2 \\rho_n\n/\\sigma_{\\mathrm{GD}}$ bits, the contraction factor of DQ-GD is the same as\nthat of unquantized GD, i.e., there is no loss due to quantization. We show\nthat no algorithm within a certain class can converge faster than\n$\\max\\{\\sigma_{\\mathrm{GD}}, 2^{-R}\\}$. Since quantizers exist with $\\rho_n \\to\n1$ as $n \\to \\infty$ (Rogers, 1963), this means that DQ-GD is asymptotically\noptimal. The principle of differential quantization continues to apply to\ngradient methods with momentum such as Nesterov's accelerated gradient descent,\nand Polyak's heavy ball method. For these algorithms as well, if the rate is\nabove a certain threshold, there is no loss in contraction factor obtained by\nthe differentially quantized algorithm compared to its unquantized counterpart.\nExperimental results on least-squares problems validate our theoretical\nanalysis.\n", " We formulate Goldbach type questions for Gaussian, Hurwitz, Octavian and\nEisenstein primes. They are different from Goldbach type statements by\nTakayoshi Mitsui from 1960 for number fields or C.A. Holben and James Jordan\nfrom 1968 for Gaussian integers. Here is what we meeasure: 1) Every even\nGaussian integer a+ib satisfying a>2, b>2 is a sum of two Gaussian primes with\npositive coefficients. 2) Every Eisenstein integer a+bw with a>3,b>3 and\nw=(1+sqrt(-3))/2 is the sum of two Eisenstein primes with positive\ncoefficients. Note that no evenness condition is asked in the Eisenstein case.\n3) Every Lipschitz integer quaternion with positive entries is the sum of two\nHurwitz primes. 4) There exists a constant K such that every Octavian integer\nwith coefficients larger than K is the sum of two Octavian primes. Except in\nthe Octonion case, where the fewest experiments were done, the statements can\nbe linked to difficult questions like Landau or Bunyakovsky conjectures. We\ntherefore look also more closely and numerically at some Hardy-Littlewood\nconstants following early computations from Daniel Shanks and Marvin Wunderlich\nfrom the 50ies and 70ies.\n", " This paper is devoted to stability results for the Gaussian logarithmic\nSobolev inequality, with explicit stability constants.\n", " One of the focus areas of modern scientific research is to reveal mysteries\nrelated to genes and their interactions. The dynamic interactions between genes\ncan be encoded into a gene regulatory network (GRN), which can be used to gain\nunderstanding on the genetic mechanisms behind observable phenotypes. GRN\ninference from time series data has recently been a focus area of systems\nbiology. Due to low sampling frequency of the data, this is a notoriously\ndifficult problem. We tackle the challenge by introducing the so-called\ncontinuous-time Gaussian process dynamical model, based on Gaussian process\nframework that has gained popularity in nonlinear regression problems arising\nin machine learning. The model dynamics are governed by a stochastic\ndifferential equation, where the dynamics function is modelled as a Gaussian\nprocess. We prove the existence and uniqueness of solutions of the stochastic\ndifferential equation. We derive the probability distribution for the Euler\ndiscretised trajectories and establish the convergence of the discretisation.\nWe develop a GRN inference method called BINGO, based on the developed\nframework. BINGO is based on MCMC sampling of trajectories of the GPDM and\nestimating the hyperparameters of the covariance function of the Gaussian\nprocess. Using benchmark data examples, we show that BINGO is superior in\ndealing with poor time resolution and it is computationally feasible.\n", " The key-agreement problem (finding a private key to use for secret messages,\notherwise referred to as the public-key distribution problem), was introduced\nby Diffie and Hellman in 1976. An approach to structuring key-agreement\nprotocols via the use of one-way associative functions was proposed in 1993 by\nRabi and Sherman. We propose here a provably strong associative one-way\nfunction based upon knot composition (answering an open problem proposed by\nRabi and Sherman whether any such associative one-way functions exist). We also\nintroduce and solve a game, exploring its relation to problems in graph and\nbraid theory and develop a new technique for computing whether a graph is\nn-colorable. En route we look at estimator and prediction problems raised in\nClassical Probability Theory using Urn problems.\n", " We present an example of a complete and minimal Gabor system consisting of\ntime-frequency shifts of a Gaussian, localized at the coordinate axes in the\ntime-frequency plane (phase space). Asymptotically, the number of\ntime-frequency shifts contained in a disk centered at the origin is only 2/pi\ntimes the number of points from the von Neumann lattice found in the same disk.\nRequiring a certain regular distribution in phase space, we show that our\nsystem has minimal density among all complete and minimal systems of\ntime-frequency shifts of a Gaussian.\n", " We study the probability that a real stationary Gaussian process has at least\n$\\eta T$ zeros in $[0,T]$ (overcrowding), or at most this number\n(undercrowding). We show that if the spectral measure of the process is\nsupported on $\\pm[B,A]$, overcrowding probability transitions from exponential\ndecay to Gaussian decay at $\\eta=\\tfrac{A}{\\pi}$, while undercrowding\nprobability undergoes the reverse transition at $\\eta=\\tfrac{B}{\\pi}$.\n", " We embark in a program of studying the problem of better approximating\nsurfaces by triangulations(triangular meshes) by considering the approximating\ntriangulations as finite metric spaces and the target smooth surface as their\nHaussdorff-Gromov limit. This allows us to define in a more natural way the\nrelevant elements, constants and invariants s.a. principal directions and\nprincipal values, Gaussian and Mean curvature, etc. By a \"natural way\" we mean\nan intrinsic, discrete, metric definitions as opposed to approximating or\nparaphrasing the differentiable notions. In this way we hope to circumvent\ncomputational errors and, indeed, conceptual ones, that are often inherent to\nthe classical, \"numerical\" approach. In this first study we consider the\nproblem of determining the Gaussian curvature of a polyhedral surface, by using\nthe {\\em embedding curvature} in the sense of Wald (and Menger). We present two\nmodalities of employing these definitions for the computation of Gaussian\ncurvature.\n", " This paper describes the concepts of Universal/ Integrally Strictly Positive\nDefinite/ $C_{0}$-Universal for the Gaussian kernel on a Hilbert space. As a\nconsequence we obtain a similar characterization for an important family of\nkernels studied and developed by Schoenberg and also on a family of\nspatial-time kernels popular on geostatistics, the Gneiting class, and its\ngeneralizations. Either by using similar techniques, or by a direct consequence\nof the Gaussian kernel on Hilbert spaces, we characterize the same concepts for\na family of kernels defined on a Hyperbolic space.\n", " In this paper we study higher order Riesz transforms associated with the\ninverse Gaussian measure given by $\\pi ^{n/2}e^{|x|^2}dx$ on $\\mathbb{R}^n$. We\nestablish $L^p(\\mathbb{R}^n,e^{|x|^2}dx)$-boundedness properties and obtain\nrepresentations as principal values singular integrals for the higher order\nRiesz transforms. New characterizations of the Banach spaces having the UMD\nproperty by means of the Riesz transforms and imaginary powers of the operator\ninvolved in the inverse Gaussian setting are given.\n", " We give a full description of complete interpolating sequences for the\nshift-invariant space generated by the Gaussian. As a consequence, we rederive\nthe known density conditions for sampling and interpolation.\n", " In the present paper we prove that the $3$-fold projective tensor product of\n$c_0$, $c_{0} \\widehat{\\otimes}_\\pi c_{0}\\widehat{\\otimes}_\\pi c_{0}$, is not\nisomorphic to a subspace of $c_{0} \\widehat{\\otimes}_\\pi c_{0}$. In particular,\nthis settles the long-standing open problem of whether $c_{0}\n\\widehat{\\otimes}_\\pi c_{0}$ is isomorphic to $c_{0} \\widehat{\\otimes}_\\pi\nc_{0}\\widehat{\\otimes}_\\pi c_{0}$. The origin of this problem goes back to Joe\nDiestel who mentioned it in a private communication to the authors of paper\n\"Unexpected subspaces of tensor products\" published in 2006.\n", " It is well known that Gaussian polynomials (i.e., $q$-binomials) describe the\ndistribution of the $area$ statistic on monotone paths in a rectangular grid.\nWe introduce two new statistics, $corners$ and $cindex$; attach ``ornaments''\nto the grid; and re-evaluate these statistics, in order to argue that all\nscrambled versions of the $cindex$ statistic are equidistributed with $area$.\nOur main result is a representation of the generating function for the\nbi-statistic $(cindex,corners)$ as a two-variable Vandermonde convolution of\nthe original Gaussian polynomial. The proof relies on explicit bijections\nbetween differently ornated paths.\n", " We show how the combined use of the generating function method and of the\ntheory of multivariable Hermite polynomials is naturally suited to evaluate\nintegrals of gaussian functions and of multiple products of Hermite\npolynomials.\n", " A theoretical model is presented which explains the dominant decoherence\nprocess in a microcavity polariton condensate. The mechanism which is invoked\nis the effect of self-phase modulation, whereby interactions transform\npolariton number fluctuations into random energy variations. The model shows\nthat the phase coherence decay, g1(t), has a Kubo form, which can be Gaussian\nor exponential, depending on whether the number fluctuations are slow or fast.\nThis fluctuation rate also determines the decay time of the intensity\ncorrelation function, g2(t), so it can be directly determined experimentally.\nThe model explains recent experimental measurements of a relatively fast\nGaussian decay for g1(t), but also predicts a regime, further above threshold,\nwhere the decay is much slower.\n", " Implicit neural representations (INRs) recently achieved great success in\nimage representation and compression, offering high visual quality and fast\nrendering speeds with 10-1000 FPS, assuming sufficient GPU resources are\navailable. However, this requirement often hinders their use on low-end devices\nwith limited memory. In response, we propose a groundbreaking paradigm of image\nrepresentation and compression by 2D Gaussian Splatting, named GaussianImage.\nWe first introduce 2D Gaussian to represent the image, where each Gaussian has\n8 parameters including position, covariance and color. Subsequently, we unveil\na novel rendering algorithm based on accumulated summation. Remarkably, our\nmethod with a minimum of 3$\\times$ lower GPU memory usage and 5$\\times$ faster\nfitting time not only rivals INRs (e.g., WIRE, I-NGP) in representation\nperformance, but also delivers a faster rendering speed of 1500-2000 FPS\nregardless of parameter size. Furthermore, we integrate existing vector\nquantization technique to build an image codec. Experimental results\ndemonstrate that our codec attains rate-distortion performance comparable to\ncompression-based INRs such as COIN and COIN++, while facilitating decoding\nspeeds of approximately 1000 FPS. Additionally, preliminary proof of concept\nshows that our codec surpasses COIN and COIN++ in performance when using\npartial bits-back coding. Code will be available at\nhttps://github.com/Xinjie-Q/GaussianImage.\n", " Many processes in chemistry, physics, and biology depend on thermally\nactivated events in which the system changes its state by surmounting an\nactivation barrier. Examples range from chemical reactions, protein folding,\nand nucleation events. Parameterized forms of the mean-field potential are\noften employed in the stochastic modeling of activated processes. In this\ncontribution, we explore the alternative of employing parameterized forms of\nthe equilibrium distribution by means of the symmetric linear combination of\ntwo gaussian functions. Such a procedure leads to flexible and convenient\nmodels for the landscape and the energy barrier whose features are controlled\nby the second moments of the gaussian functions. The rate constants are\nexamined through the solution of the corresponding diffusion problem, that is\nthe Fokker-Planck-Smoluchowski equation specified according to the\nparameterized equilibrium distribution. The numerical calculations clearly show\nthat the asymptotic limit of large barriers does not agree with the results of\nthe Kramers theory. The underlying reason is that the linear scaling of the\npotential, the procedure justifying the Kramers theory, cannot be applied when\ndealing with parameterized forms of the equilibrium distribution. A different\nkind of asymptotic analysis is then required and we introduce the appropriate\ntheory when the equilibrium distribution is represented as a symmetric linear\ncombination of two gaussian functions, first in the one-dimensional case and\nafterward in the multi-dimensional diffusion model.\n", " We give a short proof of the Brascamp-Lieb theorem, which asserts that a\ncertain general form of Young's convolution inequality is saturated by Gaussian\nfunctions. The argument is inspired by Borell's stochastic proof of the\nPr\\'ekopa-Leindler inequality and applies also to the reversed Brascamp-Lieb\ninequality, due to Barthe.\n", " In this paper, we show that a Gaussian averaged neutrino oscillation model is\nequivalent to a neutrino decoherence model. Without loss of generality, the\nanalysis is performed with two neutrino flavors. We also estimate the damping\n(or decoherence) parameter for atmospheric neutrinos and compare it to earlier\nobtained results.\n", " We introduce an unsupervised clustering algorithm to improve training\nefficiency and accuracy in predicting energies using molecular-orbital-based\nmachine learning (MOB-ML). This work determines clusters via the Gaussian\nmixture model (GMM) in an entirely automatic manner and simplifies an earlier\nsupervised clustering approach [J. Chem. Theory Comput., 15, 6668 (2019)] by\neliminating both the necessity for user-specified parameters and the training\nof an additional classifier. Unsupervised clustering results from GMM have the\nadvantage of accurately reproducing chemically intuitive groupings of frontier\nmolecular orbitals and having improved performance with an increasing number of\ntraining examples. The resulting clusters from supervised or unsupervised\nclustering is further combined with scalable Gaussian process regression (GPR)\nor linear regression (LR) to learn molecular energies accurately by generating\na local regression model in each cluster. Among all four combinations of\nregressors and clustering methods, GMM combined with scalable exact Gaussian\nprocess regression (GMM/GPR) is the most efficient training protocol for\nMOB-ML. The numerical tests of molecular energy learning on thermalized\ndatasets of drug-like molecules demonstrate the improved accuracy,\ntransferability, and learning efficiency of GMM/GPR over not only other\ntraining protocols for MOB-ML, i.e., supervised regression-clustering combined\nwith GPR(RC/GPR) and GPR without clustering. GMM/GPR also provide the best\nmolecular energy predictions compared with the ones from literature on the same\nbenchmark datasets. With a lower scaling, GMM/GPR has a 10.4-fold speedup in\nwall-clock training time compared with scalable exact GPR with a training size\nof 6500 QM7b-T molecules.\n", " We study the weighted Poincar\\'e constant $C(p,w)$ of a probability density\n$p$ with weight function $w$ using integration methods inspired by Stein's\nmethod. We obtain a new version of the Chen-Wang variational formula which, as\na byproduct, yields simple upper and lower bounds on $C(p,w)$ in terms of the\nso-called Stein kernel of $p$. We also iterate these variational formulas so as\nto build sequences of nested intervals containing the Poincar\\'e constant,\nsequences of functions converging to said constant, as well as sequences of\nfunctions converging to the solutions of the corresponding spectral problem.\nOur results rely on the properties of a pseudo inverse operator of the\nclassical Sturm-Liouville operator. We illustrate our methods on a variety of\nexamples: Gaussian functionals, weighted Gaussian, beta, gamma, Subbotin, and\nWeibull distributions.\n", " For a family of quasi-arithmetic means satisfying certain smoothness\ncondition we majorize the speed of convergence of the iterative sequence of\nself-mappings having a mean on each entry, described in the definition of\nGaussian product, to relevant mean-type mapping. We apply this result to\napproximate any continuous function which is invariant with respect to such a\nself-mappings.\n", " For a family of near banded Toeplitz matrices, generalized characteristic\npolynomials are shown to be orthogonal polynomials of two variables, which\ninclude the Chebyshev polynomials of the second kind on the deltoid as a\nspecial case. These orthogonal polynomials possess maximal number of real\ncommon zeros, which generate a family of Gaussian cubature rules in two\nvariables.\n", " In this paper we lay special stress on analyzing the topological properties\nof the lattice systems and try to ovoid the conventional ways to calculate the\ncritical points. Only those clusters with finite sizes can execute the self\nsimilar transformations of infinite hierarchies. Each ordered cluster has\nfractal dimension, their minimum relates to the edge fixed point, which accords\nwith the transformations fixed point relating to a critical point. There are\ntwo classes of systems and clusters by their connectivity. Using mathematic\nmapping method we set up cluster-spin Gaussian model solved accurately. There\nare single state and -fold coupling state in a reducible cluster, each of which\ncorresponds to a subsystem described by a Gaussian model. By the minimal\nfractal dimension a final expression of the critical points is obtained. The\ncritical points of three lattice systems are calculated, our method makes the\ncalculations very simplified and highly accurate. A possibility of existence of\ndifferent clusters at the critical temperature is discussed.\n", " Caffarelli's contraction theorem states that probability measures with\nuniformly logconcave densities on R d can be realized as the image of a\nstandard Gaussian measure by a globally Lipschitz transport map. We discuss\nsome counterexamples and obstructions that prevent a similar result from\nholding on the half-sphere endowed with a uniform measure, answering a question\nof Beck and Jerison.\n", " In this paper, we study a stochastic disclosure control problem using\ninformation-theoretic methods. The useful data to be disclosed depend on\nprivate data that should be protected. Thus, we design a privacy mechanism to\nproduce new data which maximizes the disclosed information about the useful\ndata under a strong $\\chi^2$-privacy criterion. For sufficiently small leakage,\nthe privacy mechanism design problem can be geometrically studied in the space\nof probability distributions by a local approximation of the mutual\ninformation. By using methods from Euclidean information geometry, the original\nhighly challenging optimization problem can be reduced to a problem of finding\nthe principal right-singular vector of a matrix, which characterizes the\noptimal privacy mechanism. In two extensions we first consider a scenario where\nan adversary receives a noisy version of the user's message and then we look\nfor a mechanism which finds $U$ based on observing $X$, maximizing the mutual\ninformation between $U$ and $Y$ while satisfying the privacy criterion on $U$\nand $Z$ under the Markov chain $(Z,Y)-X-U$.\n", " In this paper, we propose using curvatures in digital space for 3D object\nanalysis and recognition. Since direct adjacency has only six types of digital\nsurface points in local configurations, it is easy to determine and classify\nthe discrete curvatures for every point on the boundary of a 3D object. Unlike\nthe boundary simplicial decomposition (triangulation), the curvature can take\nany real value. It sometimes makes difficulties to find a right value for\nthreshold. This paper focuses on the global properties of categorizing\ncurvatures for small regions. We use both digital Gaussian curvatures and\ndigital mean curvatures to 3D shapes. This paper proposes a multi-scale method\nfor 3D object analysis and a vector method for 3D similarity classification. We\nuse these methods for face recognition and shape classification. We have found\nthat the Gaussian curvatures mainly describe the global features and average\ncharacteristics such as the five regions of a human face. However, mean\ncurvatures can be used to find local features and extreme points such as nose\nin 3D facial data.\n", " T. Kaluza has given a criterion for the signs of the power series of a\nfunction that is the reciprocal of another power series. In this note the\nsharpness of this condition is explored and various examples in terms of the\nGaussian hypergeometric series are given. A criterion for the monotonicity of\nthe quotient of two power series due to M. Biernacki and J. Krzy\\.z is applied.\n", " We give a polynomial time approximation scheme (PTAS) for computing the\nsupremum of a Gaussian process. That is, given a finite set of vectors\n$V\\subseteq\\mathbb{R}^d$, we compute a $(1+\\varepsilon)$-factor approximation\nto $\\mathop {\\mathbb{E}}_{X\\leftarrow\\mathcal{N}^d}[\\sup_{v\\in V}|\\langle\nv,X\\rangle|]$ deterministically in time $\\operatorname\n{poly}(d)\\cdot|V|^{O_{\\varepsilon}(1)}$. Previously, only a constant factor\ndeterministic polynomial time approximation algorithm was known due to the work\nof Ding, Lee and Peres [Ann. of Math. (2) 175 (2012) 1409-1471]. This answers\nan open question of Lee (2010) and Ding [Ann. Probab. 42 (2014) 464-496]. The\nstudy of supremum of Gaussian processes is of considerable importance in\nprobability with applications in functional analysis, convex geometry, and in\nlight of the recent breakthrough work of Ding, Lee and Peres [Ann. of Math. (2)\n175 (2012) 1409-1471], to random walks on finite graphs. As such our result\ncould be of use elsewhere. In particular, combining with the work of Ding [Ann.\nProbab. 42 (2014) 464-496], our result yields a PTAS for computing the cover\ntime of bounded-degree graphs. Previously, such algorithms were known only for\ntrees. Along the way, we also give an explicit oblivious estimator for\nsemi-norms in Gaussian space with optimal query complexity. Our algorithm and\nits analysis are elementary in nature, using two classical comparison\ninequalities, Slepian's lemma and Kanter's lemma.\n", " The quantum chaos in the finite-temperature Yang-Mills-Higgs system is\nstudied. The energy spectrum of a spatially homogeneous SU(2) Yang-Mills-Higgs\nis calculated within thermofield dynamics. Level statistics of the spectra is\nstudied by plotting nearest-level spacing distribution histograms. It is found\nthat finite temperature effects lead to a strengthening of chaotic effects,\ni.e. spectrum which has Poissonian distribution at zero temperature has\nGaussian distribution at finite-temperature.\n", " For fixed functions $G,H:[0,\\infty)\\to[0,\\infty)$, consider the rotationally\ninvariant probability density on $\\mathbb{R}^n$ of the form \\[ \\mu^n(ds) =\n\\frac{1}{Z_n} G(\\|s\\|_2)\\, e^{ - n H( \\|s\\|_2)} ds. \\] We show that when $n$ is\nlarge, the Euclidean norm $\\|Y^n\\|_2$ of a random vector $Y^n$ distributed\naccording to $\\mu^n$ satisfies a Gaussian thin-shell property: the distribution\nof $\\|Y^n\\|_2$ concentrates around a certain value $s_0$, and the fluctuations\nof $\\|Y^n\\|_2$ are approximately Gaussian with the order $1/\\sqrt{n}$. We apply\nthis thin shell property to the study of rotationally invariant random\nsimplices, simplices whose vertices consist of the origin as well as\nindependent random vectors $Y_1^n,\\ldots,Y_p^n$ distributed according to\n$\\mu^n$. We show that the logarithmic volume of the resulting simplex exhibits\nhighly Gaussian behavior, providing a generalizing and unifying setting for the\nobjects considered in Grote-Kabluchko-Th\\\"ale [Limit theorems for random\nsimplices in high dimensions, ALEA, Lat. Am. J. Probab. Math. Stat. 16,\n141--177 (2019)]. Finally, by relating the volumes of random simplices to\nrandom determinants, we show that if $A^n$ is an $n \\times n$ random matrix\nwhose entries are independent standard Gaussian random variables, then there\nare explicit constants $c_0,c_1\\in(0,\\infty)$ and an absolute constant\n$C\\in(0,\\infty)$ such that \\[\\sup_{ s \\in \\mathbb{R}} \\left| \\mathbb{P} \\left[\n\\frac{ \\log \\mathrm{det}(A^n) - \\log(n-1)! - c_0 }{ \\sqrt{ \\frac{1}{2} \\log n +\nc_1 }} < s \\right] - \\int_{-\\infty}^s \\frac{e^{ - u^2/2} du}{ \\sqrt{ 2 \\pi }}\n\\right| < \\frac{C}{\\log^{3/2}n}, \\] sharpening the $1/\\log^{1/3 + o(1)}n$ bound\nin Nguyen and Vu [Random matrices: Law of the determinant, Ann. Probab. 42 (1)\n(2014), 146--167].\n", " Let $f$ be a zero-mean continuous stationary Gaussian process on ${\\mathbb\nR}$ whose spectral measure vanishes in a $\\delta$-neighborhood of the origin.\nThen the probability that $f$ stays non-negative on an interval of length $L$\nis at most $e^{-c\\delta^2 L^2}$ with some absolute $c>0$ and the result is\nsharp without additional assumptions.\n", " Federated Learning (FL) bridges the gap between collaborative machine\nlearning and preserving data privacy. To sustain the long-term operation of an\nFL ecosystem, it is important to attract high quality data owners with\nappropriate incentive schemes. As an important building block of such incentive\nschemes, it is essential to fairly evaluate participants' contribution to the\nperformance of the final FL model without exposing their private data. Shapley\nValue (SV)-based techniques have been widely adopted to provide fair evaluation\nof FL participant contributions. However, existing approaches incur significant\ncomputation costs, making them difficult to apply in practice. In this paper,\nwe propose the Guided Truncation Gradient Shapley (GTG-Shapley) approach to\naddress this challenge. It reconstructs FL models from gradient updates for SV\ncalculation instead of repeatedly training with different combinations of FL\nparticipants. In addition, we design a guided Monte Carlo sampling approach\ncombined with within-round and between-round truncation to further reduce the\nnumber of model reconstructions and evaluations required, through extensive\nexperiments under diverse realistic data distribution settings. The results\ndemonstrate that GTG-Shapley can closely approximate actual Shapley values,\nwhile significantly increasing computational efficiency compared to the state\nof the art, especially under non-i.i.d. settings.\n", " We consider a General Relativistic generalized RW's metric,and find a field\nof Universal rotational global centripetal acceleration, numerically coincident\nwith the value of the Pioneers Anomalous one.Related subjects are also\ntreated.The rotation defined here is different from older frameworks, because\nwe propose a Gaussian metric, whose tri-space rotates relative to the time\northogonal axis, globally.The present solution is \"Machian\", paralleling our\nprevious semi-relativistic rotating solution (Berman,2007).\n", " We study the Gaussian approximation to the quantum fluctuations of the metric\nof the four dimensional anti-De Sitter spacetime. The associated massless\nscalar field has a quartic self interaction, for which we construct the\ngenerating functional of the $n$-point correlation functions. The concomitant\ninfrared divergence is cured by a mass renormalization provided by the\ncosmological constant, which is also responsible for the renormalization of the\ncoupling constant of the field theory.\n", " Aggregate time-series data like traffic flow and site occupancy repeatedly\nsample statistics from a population across time. Such data can be profoundly\nuseful for understanding trends within a given population, but also pose a\nsignificant privacy risk, potentially revealing e.g., who spends time where.\nProducing a private version of a time-series satisfying the standard definition\nof Differential Privacy (DP) is challenging due to the large influence a single\nparticipant can have on the sequence: if an individual can contribute to each\ntime step, the amount of additive noise needed to satisfy privacy increases\nlinearly with the number of time steps sampled. As such, if a signal spans a\nlong duration or is oversampled, an excessive amount of noise must be added,\ndrowning out underlying trends. However, in many applications an individual\nrealistically cannot participate at every time step. When this is the case, we\nobserve that the influence of a single participant (sensitivity) can be reduced\nby subsampling and/or filtering in time, while still meeting privacy\nrequirements. Using a novel analysis, we show this significant reduction in\nsensitivity and propose a corresponding class of privacy mechanisms. We\ndemonstrate the utility benefits of these techniques empirically with\nreal-world and synthetic time-series data.\n", " We present phaseless auxiliary-field (AF) quantum Monte Carlo (QMC)\ncalculations of the ground states of some hydrogen-bonded systems. These\nsystems were selected to test and benchmark different aspects of the new\nphaseless AF QMC method. They include the transition state of H+H$_2$ near the\nequilibrium geometry and in the van der Walls limit, as well as H$_2$O, OH, and\nH$_2$O$_2$ molecules. Most of these systems present significant challenges for\ntraditional independent-particle electronic structure approaches, and many also\nhave exact results available. The phaseless AF QMC method is used either with a\nplanewave basis with pseudopotentials or with all-electron Gaussian basis sets.\nFor some systems, calculations are done with both to compare and characterize\nthe performance of AF QMC under different basis sets and different\nHubbard-Stratonovich decompositions. Excellent results are obtained using as\ninput single Slater determinant wave functions taken from independent-particle\ncalculations. Comparisons of the Gaussian based AF QMC results with exact full\nconfiguration show that the errors from controlling the phase problem with the\nphaseless approximation are small. At the large basis-size limit, the AF QMC\nresults using both types of basis sets are in good agreement with each other\nand with experimental values.\n", " q-Gaussians are probability distributions having their origin in the\nframework of Tsallis statistics. A continuous real parameter q is\ncharacterizing them so that, in the range 1 < q < 3, the q-functions pass from\nthe usual Gaussian form, for q close to 1, to that of a heavy tailed\ndistribution, at q close to 3. The value q=2 corresponds to the\nCauchy-Lorentzian distribution. This behavior of q-Gaussian functions could be\ninteresting for a specific application, that regarding the analysis of Raman\nspectra, where Lorentzian and Gaussian profiles are the line shapes most used\nto fit the spectral bands. Therefore, we will propose q-Gaussians with the aim\nof comparing the resulting fit analysis with data available in literature. As\nit will be clear from the discussion, this is a very sensitive issue. We will\nalso provide a detailed discussion about Voigt and pseudo-Voigt functions and\ntheir role in the line shape modeling of Raman bands. We will show a\nsuccessfully comparison of these functions with q-Gaussians. The role of\nq-Gaussians in EPR spectroscopy (Howarth et al., 2003), where the q-Gaussian is\ngiven as the ''Tsallis lineshape function'', will be reported. Two examples of\nfitting Raman D and G bands with q-Gaussians are proposed too.\n", " A new proof is given for the fact that centered gaussian functions saturate\nthe Euclidean forward-reverse Brascamp-Lieb inequalities, extending the\nBrascamp-Lieb and Barthe theorems. A duality principle for best constants is\nalso developed, which generalizes the fact that the best constants in the\nBrascamp-Lieb and Barthe inequalities are equal. Finally, as the title hints,\nthe main results concerning finiteness, structure and gaussian-extremizability\nfor the Brascamp-Lieb inequality due to Bennett, Carbery, Christ and Tao are\ngeneralized to the setting of the forward-reverse Brascamp-Lieb inequality.\n", " Krein-de Branges spectral theory establishes a correspondence between the\nclass of differential operators called canonical Hamiltonian systems and\nmeasures on the real line with finite Poisson integral. We further develop this\narea by giving a description of canonical Hamiltonian systems whose spectral\nmeasures have logarithmic integral converging over the real line. This result\ncan be viewed as a spectral version of the classical Szego theorem in the\ntheory of polynomials orthogonal on the unit circle. It extends Krein-Wiener\ncompleteness theorem, a key fact in the prediction of stationary Gaussian\nprocesses.\n", " In this paper we study height fluctuations of random lozenge tilings of\npolygonal domains on the triangular lattice through nonintersecting Bernoulli\nrandom walks. For a large class of polygons which have exactly one horizontal\nupper boundary edge, we show that these random height functions converge to a\nGaussian Free Field as predicted by Kenyon and Okounkov [28]. A key ingredient\nof our proof is a dynamical version of the discrete loop equations as\nintroduced by Borodin, Guionnet and Gorin [5], which might be of independent\ninterest.\n", " Dynamical systems in nature exhibit selfsimilar fractal fluctuations and the\ncorresponding power spectra follow inverse power law form signifying long-range\nspace-time correlations identified as self-organized criticality. The physics\nof self-organized criticality is not yet identified. The Gaussian probability\ndistribution used widely for analysis and description of large data sets\nunderestimates the probabilities of occurrence of extreme events such as stock\nmarket crashes, earthquakes, heavy rainfall, etc. The assumptions underlying\nthe normal distribution such as fixed mean and standard deviation, independence\nof data, are not valid for real world fractal data sets exhibiting a scale-free\npower law distribution with fat tails. A general systems theory for fractals\nvisualizes the emergence of successively larger scale fluctuations to result\nfrom the space-time integration of enclosed smaller scale fluctuations. The\nmodel predicts a universal inverse power law incorporating the golden mean for\nfractal fluctuations and for the corresponding power spectra, i.e., the\nvariance spectrum represents the probabilities, a signature of quantum systems.\nFractal fluctuations therefore exhibit quantum-like chaos. The model predicted\ninverse power law is very close to the Gaussian distribution for small-scale\nfluctuations, but exhibits a fat long tail for large-scale fluctuations.\nExtensive data sets of Dow Jones index, Human DNA, Takifugu rubripes (Puffer\nfish) DNA are analysed to show that the space/time data sets are close to the\nmodel predicted power law distribution.\n", " This report provides an in-depth overview over the implications and novelty\nGeneralized Variational Inference (GVI) (Knoblauch et al., 2019) brings to Deep\nGaussian Processes (DGPs) (Damianou & Lawrence, 2013). Specifically, robustness\nto model misspecification as well as principled alternatives for uncertainty\nquantification are motivated with an information-geometric view. These\nmodifications have clear interpretations and can be implemented in less than\n100 lines of Python code. Most importantly, the corresponding empirical results\nshow that DGPs can greatly benefit from the presented enhancements.\n", " Perhaps the simplest first-principles approach to electronic structure is to\nfit the charge distribution of each orbital pair and use those fits wherever\nthey appear in the entire electron-electron (EE) interaction energy. The charge\ndistributions in quantum chemistry are typically represented as a sums over\nproducts of Gaussian orbital basis functions. If fitted, they are also\nrepresented as a sum over single-center Gaussian fitting basis functions. With\ntwo representations of the charge distributions, the proper definition of\nenergy is ambiguous. To remedy this, we require that the variation of the\nenergy with respect to a product of orbitals generates a fitted potential. This\nmakes the quantum-mechanical energy robust, i.e. corrected to first order for\nthe error made using an incomplete fitting basis. The coupled orbital and\nfitting equations are then the result of making the energy stationary with\nrespect to two independent sets of variables. We define the potentials and\nunique energies for methods based on the Hartree Fock model and variationally\nfit the full EE interaction in DFT. We compare implementations of variational\nfitting in DFT at six different levels for three different functionals. Our\ncalculations are performed on transition metal atoms, for which first-order\nCoulomb errors, due to an incomplete fitting basis sets, are significant.\nVariational first-order exchange and correlation errors have similar magnitude\nin all cases. Robust energy differences are much smaller, particularly in the\nlocal density approximation.\n", " The purpose of the present paper is to investigate some structural and\nqualitative aspects of two different perturbations of the parameters of\n$g$-fractions. In this context the concept of \\emph{gap} $g$-fractions is\nintroduced. While tail sequences of a continued fraction play a significant\nrole in the first perturbation, Schur fractions are used in the second\nperturbation of the $g$-parameters that are considered. Illustrations are\nprovided using Gaussian hypergeometric functions. Using a particular gap\n$g$-fraction, some members of the class of Pick functions are also identified.\n" ]

[ " We focus on the problem of manifold estimation: given a set of observations\nsampled close to some unknown submanifold $M$, one wants to recover information\nabout the geometry of $M$. Minimax estimators which have been proposed so far\nall depend crucially on the a priori knowledge of some parameters quantifying\nthe underlying distribution generating the sample (such as bounds on its\ndensity), whereas those quantities will be unknown in practice. Our\ncontribution to the matter is twofold: first, we introduce a one-parameter\nfamily of manifold estimators $(\\hat{M}_t)_{t\\geq 0}$ based on a localized\nversion of convex hulls, and show that for some choice of $t$, the\ncorresponding estimator is minimax on the class of models of $C^2$ manifolds\nintroduced in [Genovese et al., Manifold estimation and singular deconvolution\nunder Hausdorff loss]. Second, we propose a completely data-driven selection\nprocedure for the parameter $t$, leading to a minimax adaptive manifold\nestimator on this class of models. This selection procedure actually allows us\nto recover the Hausdorff distance between the set of observations and $M$, and\ncan therefore be used as a scale parameter in other settings, such as tangent\nspace estimation.\n", " In this paper transformations for matrix orthogonal polynomials in the real\nline are studied. The orthogonality is understood in a broad sense, and is\ngiven in terms of a nondegenerate continuous sesquilinear form, which in turn\nis determined by a quasidefinite matrix of bivariate generalized functions with\na well defined support. The discussion of the orthogonality for such a\nsesquilinear form includes, among others, matrix Hankel cases with linear\nfunctionals, general matrix Sobolev orthogonality and discrete orthogonal\npolynomials with an infinite support. The first transformation considered is\nthat of Geronimus type, two different methods are developed. A spectral one,\nbased on the spectral properties of the perturbing polynomial, and constructed\nin terms of the second kind functions. Then, using spectral techniques and\nspectral jets, Christoffel-Geronimus formulas for the transformed polynomials\nand norms are presented. For this type of transformations, the paper also\nproposes an alternative method, which does not require of spectral techniques.\nA discussion on matrix spectral linear transformations is presented, as well.\nThese transformations can be understood as rational perturbations of the matrix\nof bivariate generalized functions together with an addition of appropriate\nmasses, determined these last ones by the spectral properties of the polynomial\ndenominator. As for the Geronimus case, two techniques are applied, spectral\nand mixed spectral/nonspectral. Christoffel--Geronimus--Uvarov formulas are\nfound with both approaches and some applications are given. The transformation\ntheory is finally discussed in the context of the 2D non-Abelian Toda lattice\nand noncommutative KP hierarchies, understood as the theory of continuous\ntransformations of quasidefinite sesquilinear forms. This approach allows for\nthe finding of perturbed quasitau and Baker matrices.\n", " In this paper, the notion of dimension preserving approximation for\nreal-valued bivariate continuous functions, defined on a rectangular domain\n$\\rectangle$, has been introduced and several results, similar to well-known\nresults of bivariate constrained approximation in terms of dimension preserving\napproximants, have been established. Further, some clue for the construction of\nbivariate dimension preserving approximants, using the concept of fractal\ninterpolation functions, has been added. In the last part, some multi-valued\nfractal operators associated with bivariate $\\alpha$-fractal functions are\ndefined and studied.\n", " In this paper we deduce new characterizations for bivariate Bessel-Potential\nspaces defined on the unit square via B-spline quarklets. For that purpose in a\nfirst step we use univariate boundary adapted quarklets to describe univariate\nBessel-Potential spaces on intervals. To obtain the bivariate characterizations\na recent result of Hansen and Sickel is applied. It yields that each bivariate\nBessel-Potential space on a square can be written as an intersection of\nfunction spaces which have a tensor product structure. Hence our main result is\na characterization of bivariate Bessel-Potential spaces on squares in terms of\nquarklets that are tensor products of univariate quarklets on intervals.\n", " We study a class of optimization problems including matrix scaling, matrix\nbalancing, multidimensional array scaling, operator scaling, and tensor scaling\nthat arise frequently in theory and in practice. Some of these problems, such\nas matrix and array scaling, are convex in the Euclidean sense, but others such\nas operator scaling and tensor scaling are geodesically convex on a different\nRiemannian manifold. Trust region methods, which include box-constrained\nNewton's method, are known to produce high precision solutions very quickly for\nmatrix scaling and matrix balancing (Cohen et. al., FOCS 2017, Allen-Zhu et.\nal. FOCS 2017), and result in polynomial time algorithms for some geodesically\nconvex problems like operator scaling (Garg et. al. STOC 2018, B\\\"urgisser et.\nal. FOCS 2019). One is led to ask whether these guarantees also hold for\nmultidimensional array scaling and tensor scaling.\n We show that this is not the case by exhibiting instances with exponential\ndiameter bound: we construct polynomial-size instances of 3-dimensional array\nscaling and 3-tensor scaling whose approximate solutions all have doubly\nexponential condition number. Moreover, we study convex-geometric notions of\ncomplexity known as margin and gap, which are used to bound the running times\nof all existing optimization algorithms for such problems. We show that margin\nand gap are exponentially small for several problems including array scaling,\ntensor scaling and polynomial scaling. Our results suggest that it is\nimpossible to prove polynomial running time bounds for tensor scaling based on\ndiameter bounds alone. Therefore, our work motivates the search for analogues\nof more sophisticated algorithms, such as interior point methods, for\ngeodesically convex optimization that do not rely on polynomial diameter\nbounds.\n", " In recent years, several algorithms for system identification with neural\nstate-space models have been introduced. Most of the proposed approaches are\naimed at reducing the computational complexity of the learning problem, by\nsplitting the optimization over short sub-sequences extracted from a longer\ntraining dataset. Different sequences are then processed simultaneously within\na minibatch, taking advantage of modern parallel hardware for deep learning. An\nissue arising in these methods is the need to assign an initial state for each\nof the sub-sequences, which is required to run simulations and thus to evaluate\nthe fitting loss. In this paper, we provide insights for calibration of neural\nstate-space training algorithms based on extensive experimentation and analyses\nperformed on two recognized system identification benchmarks. Particular focus\nis given to the choice and the role of the initial state estimation. We\ndemonstrate that advanced initial state estimation techniques are really\nrequired to achieve high performance on certain classes of dynamical systems,\nwhile for asymptotically stable ones basic procedures such as zero or random\ninitialization already yield competitive performance.\n", " In some estimation problems, not all the parameters can be identified, which\nresults in singularity of the Fisher Information Matrix (FIM). The Cram\\'er-Rao\nBound (CRB), which is the inverse of the FIM, is then not defined. To\nregularize the estimation problem, one can impose constraints on the parameters\nand derive the corresponding CRBs. The correspondence between local\nidentifiability and FIM regularity is studied here. Furthermore the number of\nFIM singularities is shown to be equal to the number of independent constraints\nnecessary to have a well-defined constrained CRB and local identifiability. In\ngeneral, many sets of constraints can render the parameters identifiable,\ngiving different values for the CRB, that are not always relevant. When the\nconstraints can be chosen, we propose a constrained CRB, the pseudo-inverse of\nthe FIM, which gives, for a minimum number of constraints, the lowest bound on\nthe mean squared estimation error. These results are applied to two approaches\nto blind FIR multichannel estimation which allow identification of the channel\nup to a scale or phase factor. These two approaches correspond to deterministic\nand Gaussian models for the unknown channel inputs. The singularities of the\nFIMs and local identifiability are studied and the corresponding constrained\nCRBs are derived and interpreted.\n", " The gains afforded by cloud radio access network (C-RAN) in terms of savings\nin capital and operating expenses, flexibility, interference management and\nnetwork densification rely on the presence of high-capacity low-latency\nfronthaul connectivity between remote radio heads (RRHs) and baseband unit\n(BBU). In light of the non-uniform and limited availability of fiber optics\ncables, the bandwidth constraints on the fronthaul network call, on the one\nhand, for the development of advanced baseband compression strategies and, on\nthe other hand, for a closer investigation of the optimal functional split\nbetween RRHs and BBU. In this chapter, after a brief introduction to signal\nprocessing challenges in C-RAN, this optimal function split is studied at the\nphysical (PHY) layer as it pertains to two key baseband signal processing\nsteps, namely channel estimation in the uplink and channel encoding/ linear\nprecoding in the downlink. Joint optimization of baseband fronthaul compression\nand of baseband signal processing is tackled under different PHY functional\nsplits, whereby uplink channel estimation and downlink channel encoding/ linear\nprecoding are carried out either at the RRHs or at the BBU. The analysis, based\non information-theoretical arguments, and numerical results yields insight into\nthe configurations of network architecture and fronthaul capacities in which\ndifferent functional splits are advantageous. The treatment also emphasizes the\nversatility of deterministic and stochastic successive convex approximation\nstrategies for the optimization of C-RANs.\n", " It is known from the algebraic graph theory that if $L$ is the Laplacian\nmatrix of some tree $G$ with a vertex degree sequence $\\mathbf{d}=(d_1, ...,\nd_n)^\\top$ and $D$ is its distance matrix, then\n$LD+2I=(2\\cdot\\mathbf{1}-\\mathbf{d})\\mathbf{1}^\\top$, where $\\mathbf{1}$ is an\nall-ones column vector. We prove that if this matrix identity holds for the\nLaplacian matrix of some graph $G$ with a degree sequence $\\mathbf{d}$ and for\nsome matrix $D$, then $G$ is essentially a tree, and $D$ is its distance\nmatrix. This result immediately generalizes to weighted graphs. If the matrix\n$D$ is symmetric, the lower triangular part of this matrix identity is\nredundant and can be omitted. Therefore, the above bilinear matrix equation in\n$L$, $D$, and $\\mathbf{d}$ characterizes trees in terms of their Laplacian and\ndistance matrices. Applications to the extremal graph theory (especially, to\ntopological index optimization and to optimal tree problems) and to road\ntopology design are discussed.\n", " We consider chemical reaction networks modeled by a discrete state and\ncontinuous in time Markov process for the vector copy number of the species and\nprovide a novel particle filter method for state and parameter estimation based\non exact observation of some of the species in continuous time. The conditional\nprobability distribution of the unobserved states is shown to satisfy a system\nof differential equations with jumps. We provide a method of simulating a\nprocess that is a proxy for the vector copy number of the unobserved species\nalong with a weight. The resulting weighted Monte Carlo simulation is then used\nto compute the conditional probability distribution of the unobserved species.\nWe also show how our algorithm can be adapted for a Bayesian estimation of\nparameters and for the estimation of a past state value based on observations\nup to a future time.\n", " Scientists use mathematical modelling to understand and predict the\nproperties of complex physical systems. In highly parameterised models there\noften exist relationships between parameters over which model predictions are\nidentical, or nearly so. These are known as structural or practical\nunidentifiabilities, respectively. They are hard to diagnose and make reliable\nparameter estimation from data impossible. They furthermore imply the existence\nof an underlying model simplification. We describe a scalable method for\ndetecting unidentifiabilities, and the functional relations defining them, for\ngeneric models. This allows for model simplification, and appreciation of which\nparameters (or functions thereof) cannot be estimated from data. Our algorithm\ncan identify features such as redundant mechanisms and fast timescale\nsubsystems, as well as the regimes in which such approximations are valid. We\nbase our algorithm on a novel quantification of regional parametric\nsensitivity: multiscale sloppiness. Traditionally, the link between parametric\nsensitivity and the conditioning of the parameter estimation problem is made\nlocally, through the Fisher Information Matrix. This is valid in the regime of\ninfinitesimal measurement uncertainty. We demonstrate the duality between\nmultiscale sloppiness and the geometry of confidence regions surrounding\nparameter estimates made where measurement uncertainty is non-negligible.\nFurther theoretical relationships are provided linking multiscale sloppiness to\nthe Likelihood-ratio test. From this, we show that a local sensitivity analysis\n(as typically done) is insufficient for determining the reliability of\nparameter estimation, even with simple (non)linear systems. Our algorithm\nprovides a tractable alternative. We finally apply our methods to a\nlarge-scale, benchmark Systems Biology model of NF-$\\kappa$B, uncovering\npreviously unknown unidentifiabilities.\n", " We consider multiwinner elections in Euclidean space using the minimax\nChamberlin-Courant rule. In this setting, voters and candidates are embedded in\na $d$-dimensional Euclidean space, and the goal is to choose a committee of $k$\ncandidates so that the rank of any voter's most preferred candidate in the\ncommittee is minimized. (The problem is also equivalent to the ordinal version\nof the classical $k$-center problem.) We show that the problem is NP-hard in\nany dimension $d \\geq 2$, and also provably hard to approximate. Our main\nresults are three polynomial-time approximation schemes, each of which finds a\ncommittee with provably good minimax score. In all cases, we show that our\napproximation bounds are tight or close to tight. We mainly focus on the\n$1$-Borda rule but some of our results also hold for the more general\n$r$-Borda.\n", " Tropical and idempotent analysis with their relations to the Hamilton-Jacobi\nand matrix Bellman equations are discussed. Some dequantization procedures are\nimportant in tropical and idempotent mathematics. In particular, the\nHamilton-Jacobi-Bellman equation is treated as a result of the Maslov\ndequantization applied to the Schr\\\"{o}dinger equation. This leads to a\nlinearity of the Hamilton-Jacobi-Bellman equation over tropical algebras. The\ncorrespondence principle and the superposition principle of idempotent\nmathematics are formulated and examined. The matrix Bellman equation and its\napplications to optimization problems on graphs are discussed. Universal\nalgorithms for numerical algorithms in idempotent mathematics are investigated.\nIn particular, an idempotent version of interval analysis is briefly discussed.\n", " Normal estimation on 3D point clouds is a fundamental problem in 3D vision\nand graphics. Current methods often show limited accuracy in predicting normals\nat sharp features (e.g., edges and corners) and less robustness to noise. In\nthis paper, we propose a novel normal estimation method for point clouds. It\nconsists of two phases: (a) feature encoding which learns representations of\nlocal patches, and (b) normal estimation that takes the learned representation\nas input and regresses the normal vector. We are motivated that local patches\non isotropic and anisotropic surfaces have similar or distinct normals, and\nthat separable features or representations can be learned to facilitate normal\nestimation. To realise this, we first construct triplets of local patches on 3D\npoint cloud data, and design a triplet network with a triplet loss for feature\nencoding. We then design a simple network with several MLPs and a loss function\nto regress the normal vector. Despite having a smaller network size compared to\nmost other methods, experimental results show that our method preserves sharp\nfeatures and achieves better normal estimation results on CAD-like shapes.\n", " Partially-observed Boolean dynamical systems (POBDS) are a general class of\nnonlinear models with application in estimation and control of Boolean\nprocesses based on noisy and incomplete measurements. The optimal minimum mean\nsquare error (MMSE) algorithms for POBDS state estimation, namely, the Boolean\nKalman filter (BKF) and Boolean Kalman smoother (BKS), are intractable in the\ncase of large systems, due to computational and memory requirements. To address\nthis, we propose approximate MMSE filtering and smoothing algorithms based on\nthe auxiliary particle filter (APF) method from sequential Monte-Carlo theory.\nThese algorithms are used jointly with maximum-likelihood (ML) methods for\nsimultaneous state and parameter estimation in POBDS models. In the presence of\ncontinuous parameters, ML estimation is performed using the\nexpectation-maximization (EM) algorithm; we develop for this purpose a special\nsmoother which reduces the computational complexity of the EM algorithm. The\nresulting particle-based adaptive filter is applied to a POBDS model of Boolean\ngene regulatory networks observed through noisy RNA-Seq time series data, and\nperformance is assessed through a series of numerical experiments using the\nwell-known cell cycle gene regulatory model.\n", " Stochastic gradient descent (SGD) is a powerful method for large-scale\noptimization problems in the area of machine learning, especially for a\nfinite-sum formulation with numerous variables. In recent years, mini-batch SGD\ngains great success and has become a standard technique for training deep\nneural networks fed with big amount of data. Inspired by its success in deep\nlearning, we apply the idea of SGD with batch selection of samples to a classic\noptimization problem in decision version. Given $n$ unit circles, the equal\ncircle packing problem (ECPP) asks whether there exist a feasible packing that\ncould put all the circles inside a circular container without overlapping.\nSpecifically, we propose a stochastic item descent method (SIDM) for ECPP in\nlarge scale, which randomly divides the unit circles into batches and runs\nBroyden-Fletcher-Goldfarb-Shanno (BFGS) algorithm on the corresponding batch\nfunction iteratively to speedup the calculation. We also increase the batch\nsize during the batch iterations to gain higher quality solution. Comparing to\nthe current best packing algorithms, SIDM greatly speeds up the calculation of\noptimization process and guarantees the solution quality for large scale\ninstances with up to 1500 circle items, while the baseline algorithms usually\nhandle about 300 circle items. The results indicate the highly efficiency of\nSIDM for this classic optimization problem in large scale, and show potential\nfor other large scale classic optimization problems in which gradient descent\nis used for optimization.\n", " We propose a framework for bilinear multiplier operators defined via the\n(bivariate) spectral theorem. Under this framework we prove Coifman-Meyer type\nmultiplier theorems and fractional Leibniz rules. Our theory applies to\nbilinear multipliers associated with the discrete Laplacian on $\\mathbb{Z}^d,$\ngeneral bi-radial bilinear Dunkl multipliers, and to bilinear multipliers\nassociated with the Jacobi expansions.\n", " Bilevel Optimization Programming is used to model complex and conflicting\ninteractions between agents, for example in Robust AI or Privacy-preserving AI.\nIntegrating bilevel mathematical programming within deep learning is thus an\nessential objective for the Machine Learning community. Previously proposed\napproaches only consider single-level programming. In this paper, we extend\nexisting single-level optimization programming approaches and thus propose\nDifferentiating through Bilevel Optimization Programming (BiGrad) for\nend-to-end learning of models that use Bilevel Programming as a layer. BiGrad\nhas wide applicability and can be used in modern machine learning frameworks.\nBiGrad is applicable to both continuous and combinatorial Bilevel optimization\nproblems. We describe a class of gradient estimators for the combinatorial case\nwhich reduces the requirements in terms of computation complexity; for the case\nof the continuous variable, the gradient computation takes advantage of the\npush-back approach (i.e. vector-jacobian product) for an efficient\nimplementation. Experiments show that the BiGrad successfully extends existing\nsingle-level approaches to Bilevel Programming.\n", " Movable antenna (MA) is a promising technology to improve wireless\ncommunication performance by varying the antenna position in a given finite\narea at the transceivers to create more favorable channel conditions. In this\npaper, we investigate the MA-enhanced multiple-access channel (MAC) for the\nuplink transmission from multiple users each equipped with a single MA to a\nbase station (BS) with a fixed-position antenna (FPA) array. A field-response\nbased channel model is used to characterize the multi-path channel between the\nantenna array of the BS and each user's MA with a flexible position. To\nevaluate the MAC performance gain provided by MAs, we formulate an optimization\nproblem for minimizing the total transmit power of users, subject to a\nminimum-achievable-rate requirement for each user, where the positions of MAs\nand the transmit powers of users, as well as the receive combining matrix of\nthe BS are jointly optimized. To solve this non-convex optimization problem\ninvolving intricately coupled variables, we develop two algorithms based on\nzero-forcing (ZF) and minimum mean square error (MMSE) combining methods,\nrespectively. Specifically, for each algorithm, the combining matrix of the BS\nand the total transmit power of users are expressed as a function of the MAs'\nposition vectors, which are then optimized by using the proposed\nmulti-directional descent (MDD) framework. It is shown that the proposed\nZF-based and MMSE-based MDD algorithms can converge to high-quality suboptimal\nsolutions with low computational complexities. Simulation results demonstrate\nthat the proposed solutions for MA-enhanced multiple access systems can\nsignificantly decrease the total transmit power of users as compared to\nconventional FPA systems employing antenna selection under both perfect and\nimperfect field-response information.\n", " Recently, the so-called Hermite-Hadamard inequality for (operator) convex\nfunctions with one variable has known extensive several developments by virtue\nof its nice properties and various applications. The fundamental target of this\npaper is to investigate a weighted variant of Hermite-Hadamard inequality in\nmultiple variables that extends the univariate case. As an application, we\nintroduce some weighted multivariate means extending certain bivariate means\nknown in the literature.\n", " This chapter provides a comprehensive review of fundamental concepts related\nto approximate natural orbital functionals (NOFs), emphasizing their\nsignificance in quantum chemistry and physics. Focusing on fermions, the\ndiscussion excludes considerations of finite temperature and systems with a\nvariable number of particles. The theoretical foundation for approximate NOFs\nis laid out, with a particular emphasis on functional N-representability.\nVarious two-index reconstructions for the two-particle reduced density matrix\n(2RDM) are introduced, accompanied by discussions on challenges. The analysis\ndelves deeply into NOFs grounded in electron pairing, specifically focusing on\nPNOF5, PNOF7, and the Global NOF, a more versatile approach addressing both\nstatic and dynamic electron correlation components. The extension of NOFs to\nmultiplets while conserving total spin is presented, and the availability of\nopen-source implementations like DoNOF (http://github.com/DoNOF) and its\nassociated programs is highlighted. A detailed overview of optimization\nprocedures for single-point calculations is provided. Sections on geometry\noptimization and ab initio molecular dynamics, closely connected to the\navailability of analytical gradients in NOF theory, are presented. The chapter\nconcludes with the extension of NOFs to both charged and excited states.\n", " In this paper we propose the approach for constructing partitionings of hard\nvariants of the Boolean satisfiability problem (SAT). Such partitionings can be\nused for solving corresponding SAT instances in parallel. For the same SAT\ninstance one can construct different partitionings, each of them is a set of\nsimplified versions of the original SAT instance. The effectiveness of an\narbitrary partitioning is determined by the total time of solving of all SAT\ninstances from it. We suggest the approach, based on the Monte Carlo method,\nfor estimating time of processing of an arbitrary partitioning. With each\npartitioning we associate a point in the special finite search space. The\nestimation of effectiveness of the particular partitioning is the value of\npredictive function in the corresponding point of this space. The problem of\nsearch for an effective partitioning can be formulated as a problem of\noptimization of the predictive function. We use metaheuristic algorithms\n(simulated annealing and tabu search) to move from point to point in the search\nspace. In our computational experiments we found partitionings for SAT\ninstances encoding problems of inversion of some cryptographic functions.\nSeveral of these SAT instances with realistic predicted solving time were\nsuccessfully solved on a computing cluster and in the volunteer computing\nproject SAT@home. The solving time agrees well with estimations obtained by the\nproposed method.\n", " Consider real bivariate polynomials f and g, respectively having 3 and m\nmonomial terms. We prove that for all m>=3, there are systems of the form (f,g)\nhaving exactly 2m-1 roots in the positive quadrant. Even examples with m=4\nhaving 7 positive roots were unknown before this paper, so we detail an\nexplicit example of this form. We also present an O(n^{11}) upper bound for the\nnumber of diffeotopy types of the real zero set of an n-variate polynomial with\nn+4 monomial terms.\n", " Nonuniformities in the imaging characteristics of modern image sensors are a\nprimary factor in the push to develop a pixel-level generalization of the\nphoton transfer characterization method. In this paper, we seek to develop a\nbody of theoretical results leading toward a comprehensive approach for\ntackling the biggest obstacle in the way of this goal: a means of pixel-level\nconversion gain estimation. This is accomplished by developing an estimator for\nthe reciprocal-difference of normal variances and then using this to construct\na novel estimator of the conversion gain. The first two moments of this\nestimator are derived and used to construct exact and approximate confidence\nintervals for its absolute relative bias and absolute coefficient of variation,\nrespectively. A means of approximating and computing optimal sample sizes are\nalso discussed and used to demonstrate the process of pixel-level conversion\ngain estimation for a real image sensor.\n", " In the present study, we have given a corrigendum to our paper on the\napproximation properties of bivariate $(p, q)-$Bernstein operators. Recently,\nwe \\cite{kar} have defined the bivariate $(p, q)-$Bernstein operators. Later,\nwe have aware of Acar et. al \\cite{acar} already have given some moments. In\nthis case, we have revised \\cite[Lemma 2.3]{kar}.\n", " A metric graph is a 1-dimensional stratified metric space consisting of\nvertices and edges or loops glued together. Metric graphs can be naturally used\nto represent and model data that take the form of noisy filamentary structures,\nsuch as street maps, neurons, networks of rivers and galaxies. We consider the\nstatistical problem of reconstructing the topology of a metric graph embedded\nin R^D from a random sample. We derive lower and upper bounds on the minimax\nrisk for the noiseless case and tubular noise case. The upper bound is based on\nthe reconstruction algorithm given in Aanjaneya et al. (2012).\n", " Analog beamforming is an attractive and cost-effective solution to exploit\nthe benefits of massive multiple-input-multiple-output systems, by requiring\nonly one up/down-conversion chain. However, the presence of only one chain\nimposes a significant overhead in estimating the channel state information\nrequired for beamforming, when conventional digital channel estimation (CE)\napproaches are used. As an alternative, this paper proposes a novel CE\ntechnique, called periodic analog CE (PACE), that can be performed by analog\nhardware. By avoiding digital processing, the estimation overhead is\nsignificantly lowered and does not scale with number of antennas. PACE involves\nperiodic transmission of a sinusoidal reference signal by the transmitter,\nestimation of its amplitude and phase at each receive antenna via analog\nhardware, and using these estimates for beamforming. To enable such non-trivial\noperation, two reference tone recovery techniques and a novel receiver\narchitecture for PACE are proposed and analyzed, both theoretically and via\nsimulations. Results suggest that in sparse, wide-band channels and above a\ncertain signal-to-noise ratio, PACE aided beamforming suffers only a small loss\nin beamforming gain and enjoys a much lower CE overhead, in comparison to\nconventional approaches. Benefits of using PACE aided beamforming during the\ninitial access phase are also discussed.\n", " In this paper, we introduce matrix weighted Bourgain-Morrey spaces and obtain\ntwo sufficient conditions for precompact sets in matrix weighted\nBourgain-Morrey spaces. We prove that the dyadic average operator is bounded on\nsome matrix weighted Bourgain-Morrey spaces. With this result, we obtain the\nnecessity for precompact sets in some matrix weighted Bourgain-Morrey spaces.\n", " A number of landmark existence theorems of nonlinear functional analysis\nfollow in a simple and direct way from the basic separation of convex closed\nsets in finite dimension via elementary versions of the\nKnaster-Kuratowski-Mazurkiewicz principle - which we extend to arbitrary\ntopological vector spaces - and a coincidence property for so-called von\nNeumann relations. The method avoids the use of deeper results of topological\nessence such as the Brouwer fixed point theorem or the Sperner's lemma and\nunderlines the crucial role played by convexity. It turns out that the convex\nKKM principle is equivalent to the Hahn-Banach theorem, the Markov-Kakutani\nfixed point theorem, and the Sion-von Neumann minimax principle.\n", " We show that every $0$-$1$ multiplicity matrix for a simple graph $G$ is\ngenerically realisable for $G$. In particular, every multiplicity matrix for a\npath is generically realisable. We use this result to provide several families\nof joins of graphs that are realisable by a matrix with only two distinct\neigenvalues.\n", " We explore efficient optimization of toolpaths based on multiple criteria for\nlarge instances of 3D printing problems. We first show that the minimum turn\ncost 3D printing problem is NP-hard, even when the region is a simple polygon.\nWe develop SFCDecomp, a space filling curve based decomposition framework to\nsolve large instances of 3D printing problems efficiently by solving these\noptimization subproblems independently. For the Buddha model, our framework\nbuilds toolpaths over a total of 799,716 nodes across 169 layers, and for the\nBunny model it builds toolpaths over 812,733 nodes across 360 layers. Building\non SFCDecomp, we develop a multicriteria optimization approach for toolpath\nplanning. We demonstrate the utility of our framework by maximizing or\nminimizing tool path edge overlap between adjacent layers, while jointly\nminimizing turn costs. Strength testing of a tensile test specimen printed with\ntool paths that maximize or minimize adjacent layer edge overlaps reveal\nsignificant differences in tensile strength between the two classes of prints.\n", " In this series of two papers, we will prove a natural matrix weighted $T1$\ntheorem for matrix kernelled CZOs. In the current paper, we will prove matrix\nweighted norm inequalities for matrix symbolled paraproducts via a general\nmatrix weighted Carleson embedding theorem. Along the way, we will also provide\na stopping time proof of the identification of $L^p(W)$ as a weighted\nTriebel-Lizorkin space when $W$ is a matrix A${}_p$ weight.\n", " This paper reviews the main estimation and prediction results derived in the\ncontext of functional time series, when Hilbert and Banach spaces are\nconsidered, specially, in the context of autoregressive processes of order one\n(ARH(1) and ARB(1) processes, for H and B being a Hilbert and Banach space,\nrespectively). Particularly, we pay attention to the estimation and prediction\nresults, and statistical tests, derived in both parametric and non-parametric\nframeworks. A comparative study between different ARH(1) prediction approaches\nis developed in the simulation study undertaken.\n", " We show that on the curves y=e^{t(x)} where t(x) is a fixed polynomial, there\nholds a tangential Markov inequality of exponent four. Specifically, for the\nreal interval [a,b] there is a constant C such that max_{x\\in\n[a,b]}|\\frac{d}{dx}P(x,e^{t(x)})|\\leq C(deg(P))^{4} max_{x\\in\n[a,b]}|P(x,e^{t(x)})| for all bivariate polynomials P(x,y).\n", " We present a $(1- \\varepsilon)$-approximation algorithms for maximum\ncardinality matchings in disk intersection graphs -- all with near linear\nrunning time. We also present estimation algorithm that returns $(1\\pm\n\\varepsilon)$-approximation to the size of such matchings -- this algorithms\nrun in linear time for unit disks, and $O(n \\log n)$ for general disks (as long\nas the density is relatively small).\n", " In this article, we recover singularly-perturbed linear differential systems\nfrom their turning points and reduce the rank of the singularity in the\nparameter to its minimal integer value. Our treatment is Moser-based; that is\nto say it is based on the reduction criterion introduced for linear singular\ndifferential systems by Moser. Such algorithms have proved their utility in the\nsymbolic resolution of the systems of linear functional equations, giving rise\nto the package ISOLDE, as well as in the perturbed algebraic eigenvalue\nproblem. Our algorithm, implemented in the computer algebra system Maple, paves\nthe way for efficient symbolic resolution of singularly-perturbed linear\ndifferential systems as well as further applications of Moser-based reduction\nover bivariate (differential) fields.\n", " For a matrix A_2 weight W on R^p, we introduce a new notion of\nW-Calder\\'on-Zygmund matrix kernels, following earlier work in by Isralowitz.\nWe state and prove a T1 theorem for such operators and give a representation\ntheorem in terms of dyadic W-Haar shifts and paraproducts, in the spirit of\nHyt\\\"onen's Representation Theorem. Finally, by means of a Bellman function\nargument, we give sharp bounds for such operators in terms of bounds for\nweighted matrix martingale transforms and paraproducts.\n", " Most practitioners use a variant of the Alpha-Beta algorithm, a simple\ndepth-first pro- cedure, for searching minimax trees. SSS*, with its best-first\nsearch strategy, reportedly offers the potential for more efficient search.\nHowever, the complex formulation of the al- gorithm and its alleged excessive\nmemory requirements preclude its use in practice. For two decades, the search\nefficiency of \"smart\" best-first SSS* has cast doubt on the effectiveness of\n\"dumb\" depth-first Alpha-Beta. This paper presents a simple framework for\ncalling Alpha-Beta that allows us to create a variety of algorithms, including\nSSS* and DUAL*. In effect, we formulate a best-first algorithm using\ndepth-first search. Expressed in this framework SSS* is just a special case of\nAlpha-Beta, solving all of the perceived drawbacks of the algorithm. In\npractice, Alpha-Beta variants typically evaluate less nodes than SSS*. A new\ninstance of this framework, MTD(f), out-performs SSS* and NegaScout, the\nAlpha-Beta variant of choice by practitioners.\n", " We point out an error in the proof of the main result of the paper of Tanabe\net al. (2007) concerning a parameter estimation for von Mises--Fisher\ndistributions, we correct the proof of the main result and we present a short\nalternative proof.\n", " Prediction+optimization is a common real-world paradigm where we have to\npredict problem parameters before solving the optimization problem. However,\nthe criteria by which the prediction model is trained are often inconsistent\nwith the goal of the downstream optimization problem. Recently,\ndecision-focused prediction approaches, such as SPO+ and direct optimization,\nhave been proposed to fill this gap. However, they cannot directly handle the\nsoft constraints with the $max$ operator required in many real-world\nobjectives. This paper proposes a novel analytically differentiable surrogate\nobjective framework for real-world linear and semi-definite negative quadratic\nprogramming problems with soft linear and non-negative hard constraints. This\nframework gives the theoretical bounds on constraints' multipliers, and derives\nthe closed-form solution with respect to predictive parameters and thus\ngradients for any variable in the problem. We evaluate our method in three\napplications extended with soft constraints: synthetic linear programming,\nportfolio optimization, and resource provisioning, demonstrating that our\nmethod outperforms traditional two-staged methods and other decision-focused\napproaches.\n", " The pattern of a matrix M is a (0,1)-matrix which replaces all non-zero\nentries of M with a 1. A directed graph is said to support M if its adjacency\nmatrix is the pattern of M. If M is an orthogonal matrix, then a digraph which\nsupports M must satisfy a condition known as quadrangularity. We look at\nquadrangularity in tournaments and determine for which orders quadrangular\ntournaments exist. We also look at a more restrictive necessary condition for a\ndigraph to support an orthogonal matrix, and give a construction for\ntournaments which meet this condition.\n", " We prove a Kato square root estimate with anisotropically degenerate matrix\ncoefficients. We do so by doing the harmonic analysis using an auxiliary\nRiemannian metric adapted to the operator. We also derive $L^2$-solvability\nestimates for boundary value problems for divergence form elliptic equations\nwith matrix degenerate coefficients. Main tools are chain rules and Piola\ntransformations for fields in matrix weighted $L^2$ spaces, under $W^{1,1}$\nhomeomorphism.\n", " The availability of accurate localization is critical for multi-robot\nexploration strategies; noisy or inconsistent localization causes failure in\nmeeting exploration objectives. We aim to achieve high localization accuracy\nwith contemporary exploration map belief and vice versa without needing global\nlocalization information. This paper proposes a novel simultaneous exploration\nand localization (SEAL) approach, which uses Gaussian Processes (GP)-based\ninformation fusion for maximum exploration while performing communication graph\noptimization for relative localization. Both these cross-dependent objectives\nwere integrated through the Rao-Blackwellization technique. Distributed\nlinearized convex hull optimization is used to select the next-best unexplored\nregion for distributed exploration. SEAL outperformed cutting-edge methods on\nexploration and localization performance in extensive ROS-Gazebo simulations,\nillustrating the practicality of the approach in real-world applications.\n", " We construct new classes of Parseval frames for a Hilbert space which allow\nsignal reconstruction from the absolute value of the frame coefficients. As a\nconsequence, signal reconstruction can be done without using noisy phase or its\nestimation. This verifies a longstanding conjecture of the speech processing\ncommunity.\n", " We analyze combinatorial optimization problems with ordinal, i.e.,\nnon-additive, objective functions that assign categories (like good, medium and\nbad) rather than cost coefficients to the elements of feasible solutions. We\nreview different optimality concepts for ordinal optimization problems and\ndiscuss their similarities and differences. We then focus on two prevalent\noptimality concepts that are shown to be equivalent. Our main result is a\nbijective linear transformation that transforms ordinal optimization problems\nto associated standard multi-objective optimization problems with binary cost\ncoefficients. Since this transformation preserves all properties of the\nunderlying problem, problem-specific solution methods remain applicable. A\nprominent example is dynamic programming and Bellman's principle of optimality,\nthat can be applied, e.g., to ordinal shortest path and ordinal knapsack\nproblems. We extend our results to multi-objective optimization problems that\ncombine ordinal and real-valued objective functions.\n", " In this study, a hybrid method based on an Adaptive Neuro-Fuzzy Inference\nSystem (ANFIS) and Particle Swarm Optimization (PSO) for diagnosing Liver\ndisorders (ANFIS-PSO) is introduced. This smart diagnosis method deals with a\ncombination of making an inference system and optimization process which tries\nto tune the hyper-parameters of ANFIS based on the data-set. The Liver diseases\ncharacteristics are taken from the UCI Repository of Machine Learning\nDatabases. The number of these characteristic attributes are 7, and the sample\nnumber is 354. The right diagnosis performance of the ANFIS-PSO intelligent\nmedical system for liver disease is evaluated by using classification accuracy,\nsensitivity and specificity analysis, respectively. According to the\nexperimental results, the performance of ANFIS-PSO can be more considerable\nthan traditional FIS and ANFIS without optimization phase.\n", " In this work we include, for the Carnot cycle, irreversibilities of linear\nfinite rate of heat transferences between the heat engine and its reservoirs,\nheat leak between the reservoirs and internal dissipations of the working\nfluid. A first optimization of the power output, the efficiency and ecological\nfunction of an irreversible Carnot cycle, with respect to: internal temperature\nratio, time ratio for the heat exchange and the allocation ratio of the heat\nexchangers; is performed. For the second and third optimizations, the optimum\nvalues for the time ratio and internal temperature ratio are substituted into\nthe equation of power and, then, the optimizations with respect to the cost and\neffectiveness ratio of the heat exchangers are performed. Finally, a criterion\nof partial optimization for the class of irreversible Carnot engines is herein\npresented.\n", " The (COVID-19) pandemic-induced restrictions on travel and social gatherings\nhave prompted most conference organizers to move their events online. However,\nin contrast to physical conferences, virtual conferences face a challenge in\nefficiently scheduling talks, accounting for the availability of participants\nfrom different time-zones as well as their interests in attending different\ntalks. In such settings, a natural objective for the conference organizers\nwould be to maximize some global welfare measure, such as the total expected\naudience participation across all talks. However, we show that optimizing for\nglobal welfare could result in a schedule that is unfair to the stakeholders,\ni.e., the individual utilities for participants and speakers can be highly\nunequal. To address the fairness concerns, we formally define fairness notions\nfor participants and speakers, and subsequently derive suitable fairness\nobjectives for them. We show that the welfare and fairness objectives can be in\nconflict with each other, and there is a need to maintain a balance between\nthese objective while caring for them simultaneously. Thus, we propose a joint\noptimization framework that allows conference organizers to design talk\nschedules that balance (i.e., allow trade-offs) between global welfare,\nparticipant fairness and the speaker fairness objectives. We show that the\noptimization problem can be solved using integer linear programming, and\nempirically evaluate the necessity and benefits of such joint optimization\napproach in virtual conference scheduling.\n", " Triangular mass matrices for neutrinos and their charged partners contain\nfull information on neutrino mixing in a most concise form. Although the scheme\nis general and model independent, triangular matrices are typical for reducible\nbut indecomposable representations of graded Lie algebras which, in turn, are\ncharacteristic for the standard model in noncommutative geometry. The mixing\nmatrix responsible for neutrino oscillations is worked out analytically for two\nand three lepton families. The example of two families fixes the mixing angle\nto just about what is required by the Mikheyev-Smirnov-Wolfenstein resonance\noscillation of solar neutrinos. In the case of three families we classify all\nphysically plausible choices for the neutrino mass matrix and derive\ninteresting bounds on some of the moduli of the mixing matrix.\n", " We review models and approaches for error-control in order to prevent the\nbuildup of noise when gates for digital chemical and biomolecular computing\nbased on (bio)chemical reaction processes are utilized to realize stable,\nscalable networks for information processing. Solvable rate-equation models\nillustrate several recently developed methodologies for gate-function\noptimization. We also survey future challenges and possible new research\navenues.\n" ]

[ " Analog beamforming is an attractive and cost-effective solution to exploit\nthe benefits of massive multiple-input-multiple-output systems, by requiring\nonly one up/down-conversion chain. However, the presence of only one chain\nimposes a significant overhead in estimating the channel state information\nrequired for beamforming, when conventional digital channel estimation (CE)\napproaches are used. As an alternative, this paper proposes a novel CE\ntechnique, called periodic analog CE (PACE), that can be performed by analog\nhardware. By avoiding digital processing, the estimation overhead is\nsignificantly lowered and does not scale with number of antennas. PACE involves\nperiodic transmission of a sinusoidal reference signal by the transmitter,\nestimation of its amplitude and phase at each receive antenna via analog\nhardware, and using these estimates for beamforming. To enable such non-trivial\noperation, two reference tone recovery techniques and a novel receiver\narchitecture for PACE are proposed and analyzed, both theoretically and via\nsimulations. Results suggest that in sparse, wide-band channels and above a\ncertain signal-to-noise ratio, PACE aided beamforming suffers only a small loss\nin beamforming gain and enjoys a much lower CE overhead, in comparison to\nconventional approaches. Benefits of using PACE aided beamforming during the\ninitial access phase are also discussed.\n", " The modular design of planar phased arrays arranged on orthogonal\npolygon-shaped apertures is addressed and a new method is proposed to\nsynthesize domino-tiled arrays fitting multiple, generally conflicting,\nrequirements. Starting from an analytic procedure to check the\ndomino-tileability of the aperture, two multi-objective optimization techniques\nare derived to efficiently and effectively deal with small and medium/large\narrays depending on the values of the bounds for the cardinality of the\nsolution space of the admissible clustered solutions. A set of representative\nnumerical examples is reported to assess the effectiveness of the proposed\nsynthesis approach also through full-wave simulations when considering\nnon-ideal models for the radiating elements of the array.\n", " Results of experimental studies of detector arrays S-PD\n(scintillator-photodiode) and PRD (scintillator-photoreceiving device) used for\nX-ray digital radiography have shown that there exist further possibilities to\nincrease spatial resolution of this system up to 2-3 line pairs per mm.\nTheoretical analysis and experimental studies show that the two-energy\ndetection method not only allows one to detect organics on the background of\nmetal, but also substantially increases (by 3-5 times) the detection ability of\nthe system as a whole, especially if parameters of the S-PD pair are optimized,\nin particular, when ZnSe(Te) is used in the low-energy circuit. A possibility\nto distinguish, in principle, between substances with insignificant differences\nin atomic number has been theoretically proven -- by transition to multi-energy\nradiography. 3D-imaging has been realized using S-PD detector arrays.\n", " In this work we review an effective description of the interaction of a gauge\nfield with a tower of massive vector fields by introducing a non-diagonal mass\nmatrix in a gauge invariant way. Particular cases of the method with only one\nvector resonance have been used by the author elsewhere, nevertheless in this\npaper the method is developed in a general way and and we proof its main\nfeatures for an arbitrary number of vector resonances. Additionally, we show\nhow to couple the vector resonances with fermions. We find that the method can\nbe useful in order to describe the low energy phenomenology of scenarios like\nKaluza-Klein resonances of usual gauge bosons or Technicolor vector resonances\nand detailed examples are provided.\n", " The goal of motion tomography is to recover a description of a vector flow\nfield using information on the trajectory of a sensing unit. In this paper, we\ndevelop a predictor corrector algorithm designed to recover vector flow fields\nfrom trajectory data with the use of occupation kernels developed by Rosenfeld\net al.. Specifically, we use the occupation kernels as an adaptive basis; that\nis, the trajectories defining our occupation kernels are iteratively updated to\nimprove the estimation on the next stage. Initial estimates are established,\nthen under mild assumptions, such as relatively straight trajectories,\nconvergence is proven using the Contraction Mapping Theorem. We then compare to\nthe established method by Chang et al. by defining a set of error metrics. We\nfound that for simulated data, which provides a ground truth, our method offers\na marked improvement and that for a real-world example we have similar results\nto the established method.\n", " Solving analytic systems using inversion can be implemented in a variety of\nways. One method is to use Lagrange inversion and variations. Here we present a\ndifferent approach, based on dual vector fields. For a function analytic in a\nneighborhood of the origin in the complex plane, we associate a vector field\nand its dual, an operator version of Fourier transform. The construction\nextends naturally to functions of several variables. We illustrate with various\nexamples and present an efficient algorithms readily implemented symbolically\nin Maple while suitable as well for numerical computations using languages such\nas C or Java.\n", " Using the language of Riordan arrays, we look at two related iterative\nprocesses on matrices and determine which matrices are invariant under these\nprocesses. In a special case, the invariant sequences that arise are\nconjectured to have Hankel transforms that obey Somos-4 recurrences. A notion\nof eigentriangle for a number triangle emerges and examples are given,\nincluding a construction of the Takeuchi numbers.\n", " In this paper, we derive a formula to express the maximum number of\nnon-intersecting diagonals of arbitrary length that can be drawn in n x n\nsquare arrays, where n is a multiple of l+1.\n", " We consider chemical reaction networks modeled by a discrete state and\ncontinuous in time Markov process for the vector copy number of the species and\nprovide a novel particle filter method for state and parameter estimation based\non exact observation of some of the species in continuous time. The conditional\nprobability distribution of the unobserved states is shown to satisfy a system\nof differential equations with jumps. We provide a method of simulating a\nprocess that is a proxy for the vector copy number of the unobserved species\nalong with a weight. The resulting weighted Monte Carlo simulation is then used\nto compute the conditional probability distribution of the unobserved species.\nWe also show how our algorithm can be adapted for a Bayesian estimation of\nparameters and for the estimation of a past state value based on observations\nup to a future time.\n", " The matrix element for muon capture by a proton is calculated to O(p^3)\nwithin heavy baryon chiral perturbation theory using the new O(p^3) Lagrangian\nof Ecker and Mojzis. External nucleon fields are renormalized using the\nappropriate definition of the wave function renormalization factor Z_N. Our\nexpression for Z_N differs somewhat from that found in existing literature, but\nis the one which is consistent with the Lagrangian we use and the one which\nensures, within our approach, the nonrenormalization of the vector coupling as\nrequired by the conserved vector current. Expressions for the standard muon\ncapture form factors are derived and compared to experimental data and we\ndetermine three of the coefficients of the Ecker - Mojzis Lagrangian, namely,\nb_7, b_{19}, and b_{23}.\n", " Classical orthogonal polynomials have widespread applications including in\nnumerical integration, solving differential equations, and interpolation.\nChanging basis between classical orthogonal polynomials can affect the\nconvergence, accuracy, and stability of solutions.\n We provide a general method for changing basis between any pair of classical\northogonal polynomials by using algebraic expressions called coefficient\nfunctions that evaluate to connection coefficients. The method builds directly\non previous work on the change of basis groupoid. The scope has fifteen kinds\nof classical orthogonal polynomials including the classes of Jacobi, Gegenbauer\nand generalized Laguerre polynomials.\n The method involves the mappings to and from the monomials for these\npolynomial bases. Sixteen coefficient functions appear to be new for\npolynomials that do not have definite parity. We derive the remainder from\nseven sources in the literature. We give a complete summary of thirty\ncoefficient functions.\n This enables change of basis to be defined algebraically and uniformly\nbetween any pair of classical orthogonal polynomial bases by using a vector dot\nproduct to compose two coefficient functions to give a third.\n A definition of Jacobi polynomials uses a basis of shifted monomials. We find\na key new mapping between the monomials and Jacobi polynomials by using a\ngeneral mapping between shifted monomials and monomials. It yields new mappings\nfor the Chebyshev polynomials of the third and fourth kinds and their shifted\nversions.\n", " The Standard Model for electroweak interactions derives four vector gauge\nboson from an SU(2)xU(1) symmetry. A doublet of complex scalar (Higgs) bosons\nis added to generate masses by spontaneous symmetry breaking. Both, the four\nvector bosons and four scalar bosons corresponding to the Higgs bosons of the\nStandard Model are shown to emerge as gauge bosons from an underlying SU(3)\nsymmetry of electroweak lepton interactions. For the known leptons, the latter\nsymmetry implies purely axial coupling of the Z boson to charged leptons, if\nthe photon is required to have purely vector coupling. The corresponding\nWeinberg angle is determined by sin{\\theta} = 1/2. The same SU(3) symmetry\nholds for quarks as well, if the observed fractional electric quark charges are\naverages of integer charges, according to the Han-Nambu scheme.\n", " We compute the two-loop massless QCD corrections to the helicity amplitudes\nfor the production of two massive vector bosons in quark-antiquark\nannihilation, allowing for an arbitrary virtuality of the vector bosons: $q\n\\bar q' \\to V_1V_2$. Combining with the leptonic decay currents, we obtain the\nfull two-loop QCD description of the corresponding electroweak four-lepton\nproduction processes. The calculation is performed by projecting the two-loop\ndiagrams onto an appropriate basis of Lorentz structures. All two-loop Feynman\nintegrals are reduced to a basis of master integrals, which are then computed\nusing the differential equations method and optimised for numerical\nperformance. We provide a public C++ code which allows for fast and precise\nnumerical evaluations of the amplitudes.\n", " We study the ground-state properties of electrons confined to a quantum wire\nand subject to a smoothly modulated Rashba spin-orbit coupling. When the period\nof the modulation becomes commensurate with the band filling, the Rashba\ncoupling drives a quantum phase transition to a nonmagnetic insulating state.\nUsing bosonization and a perturbative renormalization group approach, we find\nthat this state is robust against electron-electron interactions. The gaps to\ncharge- and spin excitations scale with the amplitude of the Rashba modulation\nwith a common interaction-dependent exponent. An estimate of the expected size\nof the charge gap, using data for a gated InAs heterostructure, suggests that\nthe effect can be put to practical use in a future spin transistor design.\n", " In this paper we lay special stress on analyzing the topological properties\nof the lattice systems and try to ovoid the conventional ways to calculate the\ncritical points. Only those clusters with finite sizes can execute the self\nsimilar transformations of infinite hierarchies. Each ordered cluster has\nfractal dimension, their minimum relates to the edge fixed point, which accords\nwith the transformations fixed point relating to a critical point. There are\ntwo classes of systems and clusters by their connectivity. Using mathematic\nmapping method we set up cluster-spin Gaussian model solved accurately. There\nare single state and -fold coupling state in a reducible cluster, each of which\ncorresponds to a subsystem described by a Gaussian model. By the minimal\nfractal dimension a final expression of the critical points is obtained. The\ncritical points of three lattice systems are calculated, our method makes the\ncalculations very simplified and highly accurate. A possibility of existence of\ndifferent clusters at the critical temperature is discussed.\n", " To the foundation of a principally new short-term forecasting method there\nhas been laid down a theory of surrounding us world's creation and of physical\nvacuum as a result of interaction of byuons - discrete objects. The definition\nof the byuon contains the cosmological vector-potential A_g - a novel\nfundamental vector constant. This theory predicts a new anisotropic interaction\nof nature objects with the physical vacuum. A peculiar \"tap\" to gain new energy\n(giving rise to an earthquake) are elementary particles because their masses\nare proportional to the modulus of some summary potential A_sum that contains\npotentials of all known fields. The value of A_sum cannot be larger than the\nmodulus of A_g. In accordance with the experimental results a new force\nassociated with A_sum ejects substance from the area of the weakened A_sum\nalong a conical formation with the opening of 100 +- 10 and the axis directed\nalong the vector A_sum. This vector has the following coordinates in the second\nequatorial coordinate system: right ascension alpha = 293 +- 10, declination\ndelta = 36 +- 10. Nearly 100% probability of an earthquake (earthquakes of 6\npoints strong and more by the Richter scale) arises when in the process of the\nearth rotation the zenith vector of a seismically dangerous region and/or the\nvectorial potential of Earth's magnetic fields are in a certain way oriented\nrelative to the vector A_g. In the work, basic models and standard mechanisms\nof earthquakes are briefly considered, results of processing of information on\nthe earthquakes in the context of global spatial anisotropy caused by the\nexistence of the vector A_g, are presented, and an analysis of them is given.\n", " Commencing with the composite fermion description of the $\\nu=5/2$ fractional\nquantum Hall effect, we study the dynamics of the edge neutral Majorana\nfermions. We confirm that these neutral modes are chiral and show that a\nconventional p-wave pairing interaction between CFs does not contribute to the\ndynamics of the edge neutral fermions. We find an important bilinear coupling\nbetween the charged and neutral modes. We show that owing to this coupling, the\ndispersion of the neutral modes is linear and their velocities are proportional\nto the wave vector of the charged mode. This dynamic origin of the motion of\nthe edge Majorana fermions was never expected before.\n", " Anharmonicities provide a wealth of information about the vibrational\ndynamics, mode coupling and energy transfer within a polyatomic system. In this\ncontribution we show how driven molecular dynamics trajectories can be used to\nextract anharmonicity properties under very short times of a few hundreds of\nvibrational periods, using two exciting fields at- and slightly off-resonance.\nDetailed analyses on generic quartic potential energy surfaces and applications\nto various model systems are presented, giving good agreement with perturbation\ntheory. Application to a realistic molecule, cubane (C$_8$H$_8$), modeled with\na tight-binding quantum force field, further indicates how the method can be\napplied in practical cases.\n", " The increasing cosmic ray statistics collected by present experiments and the\nfuture prospects with new large arrays demand accurate calculations of the\nextensive air shower parameters. The energy of the primary particle is\nestimated by ground arrays fitting a lateral distribution function (LDF) to the\nparticle densities at a given observing level. However, the lack of appropriate\nparameterization for these distributions, able to reproduce the data collected\nfrom all arrival directions, makes it difficult the experimental analysis.\nStarting from analytical LDF for vertical showers we study the atmospheric\ndepth dependence of the shower parameters, which is necessary to obtain the\ncorresponding particle density distributions for inclined showers. In our\napproach both geometrical and atmospheric attenuation effects producing a\nlateral asymmetry in non-vertical showers are included. The resulting electron\nand muon LDF fit very well the Monte Carlo simulated data at all zenith angles.\n", " We propose an efficient deterministic method to calculate the full\nconfiguration interaction (FCI) ground state energy. This method leverages the\nsparseness of the Lanczos basis vectors that span the Krylov subspace\nassociated with the Hamiltonian to rapidly calculate the lowest eigenvalue of\nthe effective Hamiltonian formed in this basis. By iteratively performing the\nspanning and diagonalization steps, this method is capable of rapidly reaching\nchemical accuracy for a variety of strongly correlated molecules, including\nstretched N$_2$ and C$_2$, in merely several tens of iterations. To accelerate\nour algorithm, we implement parallelized sparse matrix-sparse vector\nmultiplication, which features a novel hashing method that establishes a\none-to-one correspondence between the set of full configuration Slater\ndeterminants and a set of consecutive integers. A near-linear acceleration was\nobserved.\n", " Vector field design on surfaces was originally motivated by applications in\ngraphics such as texture synthesis and rendering. In this paper, we consider\nthe idea of vector field design with a new motivation from computational\ntopology. We are interested in designing and visualizing vector fields to aid\nthe study of Morse functions, Morse vector fields, and Morse-Smale complexes.\nTo achieve such a goal, we present MVF Designer, a new interactive design\nsystem that provides fine-grained control over vector field geometry, enables\nthe editing of vector field topology, and supports a design process in a simple\nand efficient way using elementary moves, which are actions that initiate or\nadvance our design process. Our system allows mathematicians to explore the\ncomplex configuration spaces of Morse functions, their gradients, and their\nassociated Morse-Smale complexes. Understanding these spaces will help us\nexpand further their applicability in topological data analysis and\nvisualization.\n", " Circadian clocks are biochemical oscillators that allow organisms to estimate\nthe time of the day. These oscillators are inherently noisy due to the discrete\nnature of the reactants and the stochastic character of their interactions. To\nkeep these oscillators in sync with the daily day-night rhythm in the presence\nof noise, circadian clocks must be coupled to the dark-light cycle. In this\nmanuscript, we study the entrainment of phase oscillators as a function of the\nintrinsic noise in the system. Using stochastic simulations, we compute the\noptimal coupling strength, intrinsic frequency and shape of the phase-response\ncurve, that maximize the mutual information between the phase of the clock and\ntime. We show that the optimal coupling strength and intrinsic frequency\nincrease with the noise, but that the shape of the phase-response curve varies\nnon-monotonically with the noise: in the low-noise regime, it features a\ndeadzone that increases in width as the noise increases, while in the\nhigh-noise regime, the width decreases with the noise. These results arise from\na trade-off between maximizing stability -noise suppression - and maximizing\nlinearity of the input-output, i.e. time- phase, relation. We also show that\nthree analytic approximations -the linear-noise approximation, the\nphase-averaging method, and linear-response theory- accurately describe\ndifferent regimes of the coupling strength and the noise.\n", " We investigate charge pumping in ferromagnet/triplet superconductor junctions\nwhere the magnetization of the ferromagnet is inhomogeneous and dynamical. It\nis shown that charge current is pumped due to the coupling of the localized\nspin with triplet vector spin chirality, vector spin chirality formed by the\ntriplet vector of Cooper pairing. Physical mechanism of the charge pumping is\nalso discussed.\n", " The Euclidean-time hadronic current correlation functions, $G_P(\\tau, T)$ and\n$G_V(\\tau, T)$, of pseudoscalar and vector currents have recently been\ncalculated in lattice simulations of QCD and have been used to obtain the\ncorresponding spectral functions. We have used the Nambu-Jona-Lasinio (NJL)\nmodel to calculate such spectral functions, as well as the Euclidean-time\ncorrelators, and have made a comparison to the lattice results for the\ncorrelators. We find evidence for the type of temperature dependence of the NJL\ncoupling parameters that we have used in previous studies of the mesonic\nconfinement-deconfinement transition. We also see that the spectral functions\nobtained when using the maximum-entropy-method (MEM) and the lattice data\ndiffer from the spectral functions that we calculate in our chiral model.\nHowever, our results for the Euclidean-time correlators are in general\nagreement with the lattice results, with better agreement when our\ntemperature-dependent coupling parameters are used than when\ntemperature-independent parameters are used for the NJL model. We also discuss\nsome additional evidence for the utility of temperature-dependent coupling\nparameters for the NJL model. For example, if the constituent quark mass at T=0\nis $352 {MeV}$ in the chiral limit, the transition temperature is $T_c=208\n{MeV}$ for the NJL model with a standard momentum cutoff parameter. (If a\nGaussian momentum cutoff is used, we find $T_c=225 {MeV}$ in the chiral limit,\nwith $m=368 {MeV}$ at T=0.) The introduction of a weak temperature dependence\nfor the coupling constant will move the value of $T_c$ into the range 150-170\nMeV, which is more in accord with what is found in lattice simulations of QCD\nwith dynamical quarks.\n", " Torsion propagation and torsion-spin coupling are studied in the perspective\nof the Velo-Zwanziger method of analysis; specifically, we write the most\nextensive dynamics of the torsion tensor and the most exhaustive coupling that\nis permitted between torsion and spinors, and check the compatibility with\nconstraints and hyperbolicity and causality of field equations: we find that\nsome components of torsion and many terms of the torsion-spin interaction will\nbe restricted away and as a consequence we will present the most general theory\nthat is compatible with all restrictions.\n", " It has long been conjectured that generic dynamical systems has finite\nperiodic orbits, ever since the time of Poincar\\'e. In this article, a\nperturbation method is proposed for the $C^r$ closing of periodic orbits. This\nmethod is applicable to both time-varying and time-invariant vector fields.\n", " The explosive growth of data results in more scarce spectrum resources. It is\nimportant to optimize the system performance under limited resources. In this\npaper, we investigate how to achieve weighted throughput (WTP) maximization for\ncell-free (CF) multiuser MIMO (MU-MIMO) multicarrier (MC) systems through\nresource allocation (RA), in the cases of finite blocklength (FBL) and infinite\nblocklength (INFBL) regimes. To ensure the quality of service (QoS) of each\nuser, particularly for the block error rate (BLER) and latency in the FBL\nregime, the WTP gets maximized under the constraints of total power consumption\nand required QoS metrics. Since the channels vary in different subcarriers\n(SCs) and inter-user interference strengths, the WTP can be maximized by\nscheduling the best users in each time-frequency (TF) resource and advanced\nbeamforming design, while the resources can be fully utilized. With this\nmotivation, we propose a joint user scheduling (US) and beamforming design\nalgorithm based on the successive convex approximation (SCA) and gene-aided\n(GA) algorithms, to address a mixed integer nonlinear programming (MINLP)\nproblem. Numerical results demonstrate that the proposed RA outperforms the\ncomparison schemes. And the CF system in our scenario is capable of achieving\nhigher spectral efficiency than the centralized antenna systems (CAS).\n", " Estimating causal relations is vital in understanding the complex\ninteractions in multivariate time series. Non-linear coupling of variables is\none of the major challenges inaccurate estimation of cause-effect relations. In\nthis paper, we propose to use deep autoregressive networks (DeepAR) in tandem\nwith counterfactual analysis to infer nonlinear causal relations in\nmultivariate time series. We extend the concept of Granger causality using\nprobabilistic forecasting with DeepAR. Since deep networks can neither handle\nmissing input nor out-of-distribution intervention, we propose to use the\nKnockoffs framework (Barberand Cand`es, 2015) for generating intervention\nvariables and consequently counterfactual probabilistic forecasting. Knockoff\nsamples are independent of their output given the observed variables and\nexchangeable with their counterpart variables without changing the underlying\ndistribution of the data. We test our method on synthetic as well as real-world\ntime series datasets. Overall our method outperforms the widely used vector\nautoregressive Granger causality and PCMCI in detecting nonlinear causal\ndependency in multivariate time series.\n", " We present an approach to generalized Riordan arrays which is based on\noperations in one large group of lower triangular matrices. This allows for\ndirect proofs of many properties of weighted Sheffer sequences, and shows that\nall the groups arising from different weights are isomorphic since they are\nconjugate. We also prove a result about the intersection of two generalized\nRiordan with different weights.\n", " We derive formulas for $(i)$ the number of toroidal $n\\times n$ binary\narrays, allowing rotation of rows and/or columns as well as matrix\ntransposition, and $(ii)$ the number of toroidal $n\\times n$ binary arrays,\nallowing rotation and/or reflection of rows and/or columns as well as matrix\ntransposition.\n", " We present the first sublinear memory sketch that can be queried to find the\nnearest neighbors in a dataset. Our online sketching algorithm compresses an N\nelement dataset to a sketch of size $O(N^b \\log^3 N)$ in $O(N^{(b+1)} \\log^3\nN)$ time, where $b < 1$. This sketch can correctly report the nearest neighbors\nof any query that satisfies a stability condition parameterized by $b$. We\nachieve sublinear memory performance on stable queries by combining recent\nadvances in locality sensitive hash (LSH)-based estimators, online kernel\ndensity estimation, and compressed sensing. Our theoretical results shed new\nlight on the memory-accuracy tradeoff for nearest neighbor search, and our\nsketch, which consists entirely of short integer arrays, has a variety of\nattractive features in practice. We evaluate the memory-recall tradeoff of our\nmethod on a friend recommendation task in the Google Plus social media network.\nWe obtain orders of magnitude better compression than the random projection\nbased alternative while retaining the ability to report the nearest neighbors\nof practical queries.\n", " We consider tensor-vector theories with varying the space-time-matter\ncoupling constant (varying Einstein velocity) in a spatially flat FRW universe.\nWe examine the dynamics of this model by dynamical system method assuming a\n\\Lambda CDM background and we find some exact solutions by considering the\ncharacter of critical points of the theory and their stability conditions. Then\nwe reconstruct the potential V(A^2) and the coupling Z(A^2) by demanding a\nbackground \\Lambda CDM cosmology. Also we set restrictions on the varying\nEinstein velocity to solve the horizon problem. This gives a selection rule for\nchoosing the appropriate stable solution. We will see that it is possible to\nproduce the background expansion history H(z) indicated by observations.\nFinally we will discuss the behavior of the speed of light (c_E) for those\nsolutions.\n", " We study the vector-axialvector mixing in a hot medium and its evolution\ntoward the chiral phase transition using different symmetry restoration\nscenarios based on the generalized hidden local symmetry framework. We show\nthat the presence of the $a_1$ meson reduces the vector spectral function\naround $\\rho$ meson mass and enhances it around $a_1$ meson mass. The coupling\nstrength of $a_1$ to $\\rho$ and $\\pi$ vanishes at the critical temperature due\nto the degenerate $\\rho$-$a_1$ masses. This feature holds rigorously in the\nchiral limit and still stays intact to good approximation for the physical pion\nmass.\n", " By using the background field method of QCD in a path integral approach, we\nderive the equation of motion for the classical chromofield and for the gluon\nin a system containing the gluon and the classical chromofield simultaneously.\nThis inhomogeneous field equation contains a current term, which is the\nexpectation value of a composite operator including linear, square and cubic\nterms of the gluon field. We also derive identities which the current should\nobey from the gauge invariance. We calculate the current at the leading order\nwhere the current induced by the gluon is opposite in sign to that induced by\nthe quark. This is just the feature of the non-Abelian gauge field theory which\nhas asymptotic freedom. Physically, the induced current can be treated as the\n'displacement' current in the polarized vacuum, and its effect is equivalent to\nredefining the field and the coupling constant.\n", " We consider elements of finite order in the Riordan group $\\cal R$ over a\nfield of characteristic $0$. Viewing $\\cal R$ as a semi-direct product of\ngroups of formal power series, we solve, for all $n \\geq 2$, two foundational\nquestions posed by L. Shapiro for the case $n = 2$ (`involutions'): Given a\nformal power series $F(x)$ of finite compositional order and an integer $n\\geq\n2$, Theorem 1 states, exactly which $g(x)$ make $\\big(g(x), F(x)\\big)$ a\nRiordan element of order $n$. Theorem 2 classifies finite-order Riordan group\nelements up to conjugation in $\\cal R$. Viewing $\\cal R$ as a group of infinite\nlower triangular matrices, we interpret Theorem 1 in terms of existence of\neigenvectors and Theorem 2 as a normal form for finite order Riordan arrays\nunder similarity. These lead to Theorem 3, a formula for all eigenvectors of\nfinite order Riordan arrays; and we show how this can lead to interesting\ncombinatorial identities. We then relate our work to papers of Cheon and Kim\nwhich motivated this paper and we solve the Open question which they posed.\nFinally, this circle of ideas gives a new proof of C. Marshall's theorem, which\nfinds the unique $F(x)$, given bi-invertible $g(x)$, such that $\\big(g(x),\nF(x))$ is an involution.\n", " Let J and J* be subsets of Z+ such that 0,1\\in J and 0\\in J*. For infinitely\nmany n, let k=(k_1,..., k_n) be a vector of nonnegative integers whose sum M is\neven. We find an asymptotic expression for the number of multigraphs on the\nvertex set {1,..., n} with degree sequence given by k, such that every loop has\nmultiplicity in J* and every non-loop edge has multiplicity in J. Equivalently,\nthese are symmetric integer matrices with values J* allowed on the diagonal and\nJ off the diagonal. Our expression holds when the maximum degree K satisfies K\n= o(M^(1/3)). We prove this result using the switching method, building on an\nasymptotic enumeration of simple graphs with given degrees (McKay and Wormald,\n1991). Our application of the switching method introduces a novel way of\ncombining several different switching operations into a single computation.\n", " We give a complete characterization of limiting interpolation spa\\-ces for\nthe real method of interpolation using extrapolation theory. For this purpose\nthe usual tools (e.g., Boyd indices or the boundedness of Hardy type operators)\nare not appropriate. Instead, our characterization hinges upon the boundedness\nof some simple operators (e.g. $f\\mapsto f(t^{2})/t$, or $f\\mapsto f(t^{1/2}%\n)$) acting on the underlying lattices that are used to control the $K$- and\n$J$-functionals. Reiteration formulae, extending Holmstedt's classical\nreiteration theorem to limiting spaces, are also proved and characterized in\nthis fashion. The resulting theory gives a unified roof to a large body of\nliterature that, using ad-hoc methods, had covered only special cases of the\nresults obtained here. Applications to Matsaev ideals, Grand Lebesgue spaces,\nBourgain-Brezis-Mironescu-Maz'ya-Shaposhnikova limits, as well as a new vector\nvalued extrapolation theorems, are provided.\n", " Determination of trilinear Higgs coupling\n($\\lambda_{HHH}=\\kappa\\lambda_{HHH}^{\\rm SM}$) through Higgs pair productions\nis a major motivation for the LHC high luminosity phase. We perform a detailed\ncollider simulation to explore the potential of measuring $\\lambda_{HHH}$ in\nthe $VHH$ ($V=W,Z$) production at the HL-LHC. We find that the trilinear Higgs\ncoupling in the SM ($\\lambda_{HHH}^{\\rm SM}$) could be measured at the level of\n$1.3\\sigma$. Combining with the gluon fusion, vector boson fusion and\n$t\\bar{t}HH$ channels, $\\lambda_{HHH}^{\\rm SM}$ is expected to be measured at\nthe level of $3.13\\sigma$. If no evidence of Higgs pair productions were\nobserved, the $VHH$ production, together with the gluon fusion channel, would\nimpose a bound of $0.5\\leq \\kappa \\leq 2.2$ at the 95\\% confidence level.\n", " Hard diffractive electroproduction of longitudinally polarized vector mesons\nis calculated within the leading $\\alpha_s\\ln{Q^2\\over\\Lambda_{QCD}^2}$\napproximation of QCD using the leading order parton densities within the\nnucleon. Novel QCD features of the production of excited states and of the\nrestoration of flavor symmetry are analyzed. At the onset of the asymptotic\nregime, our analysis finds an important role of quark Fermi motion within the\ndiffractively produced vector mesons, and we suggest to use this effect to\nmeasure the high momentum tail of the wave function of the vector mesons. We\ndeduce a kinematical boundary for the region of applicability of the\ndecomposition of the hard amplitudes over powers of $Q^2$ and/or a limit on the\nincrease of the cross sections of hard processes at small $x$, and briefly\nanalyze its consequences. We also estimate the nuclear attenuation of the\ndiffractive electroproduction of vector mesons and compare with estimates of\nthe shadowing of the longitudinal structure function.\n", " Computational models of decisionmaking must contend with the variance of\ncontext and any number of possible decisions that a defined strategic actor can\nmake at a given time. Relying on cognitive science theory, the authors have\ncreated an algorithm that captures the orientation of the actor towards an\nobject and arrays the possible decisions available to that actor based on their\ngiven intersubjective orientation. This algorithm, like a traditional K-means\nclustering algorithm, relies on a core-periphery structure that gives the\nlikelihood of moves as those closest to the cluster's centroid. The result is\nan algorithm that enables unsupervised classification of an array of decision\npoints belonging to an actor's present state and deeply rooted in cognitive\nscience theory.\n", " In this paper, we propose using curvatures in digital space for 3D object\nanalysis and recognition. Since direct adjacency has only six types of digital\nsurface points in local configurations, it is easy to determine and classify\nthe discrete curvatures for every point on the boundary of a 3D object. Unlike\nthe boundary simplicial decomposition (triangulation), the curvature can take\nany real value. It sometimes makes difficulties to find a right value for\nthreshold. This paper focuses on the global properties of categorizing\ncurvatures for small regions. We use both digital Gaussian curvatures and\ndigital mean curvatures to 3D shapes. This paper proposes a multi-scale method\nfor 3D object analysis and a vector method for 3D similarity classification. We\nuse these methods for face recognition and shape classification. We have found\nthat the Gaussian curvatures mainly describe the global features and average\ncharacteristics such as the five regions of a human face. However, mean\ncurvatures can be used to find local features and extreme points such as nose\nin 3D facial data.\n", " Using the kinetic approach, we study the impact of the charged particles\ndynamics due to the Schwinger effect on the electric field evolution during\ninflation. As a simple model of the electric field generation, we consider the\nkinetic coupling of the electromagnetic field to the inflaton via the term\n$f^{2}(\\phi)F_{\\mu\\nu}F^{\\mu\\nu}$ with the Ratra coupling function\n$f=\\exp(\\beta\\phi/M_{p})$. The production of charged particles is taken into\naccount in the Boltzmann kinetic equation through the Schwinger source term.\nProduced particles are thermalized due to collisions which we model by using\nthe collision integral in the self-consistent relaxation time approximation. We\nfound that the current of created particles exhibits a non-Markovian character\nand cannot be described by a simple Ohm's law relation $j\\propto E$. On the\ncontrary, the electric current as well as the electric field are oscillatory\nfunctions of time with decreasing amplitudes and a phase difference due to the\nballistic motion of charged carriers. Our qualitative results are checked by\nusing a hydrodynamic approach. Deriving a closed system of equations for the\nnumber, current, and energy densities of charged particles and determining its\nsolution, we find a good agreement with the results obtained in the kinetic\napproach.\n", " We present results for Higgs boson pair production with variations of the\ntrilinear Higgs boson self-coupling at next-to-leading order (NLO) in QCD\nincluding the full top quark mass dependence. Differential results at 14 TeV\nare presented, and we discuss the implications of anomalous trilinear couplings\nas well as differences between the PYTHIA 8.2 and HERWIG 7.1 parton showers in\ncombination with POWHEG. The implementation of the NLO QCD calculation with\nvariable Higgs boson self-coupling is made publicly available in the\nPOWHEG-BOX-V2 Monte Carlo framework. A simple method for using the new\nimplementation to study also variations of the top quark Yukawa coupling is\ndescribed.\n", " In this paper, we introduce and study the Weyl transform of functions which\nare integrable with respect to a vector measure on a phase space associated to\na locally compact abelian group. We also study the Weyl transform of vector\nmeasures. Later, we also introduce and study the convolution of functions from\n$L^p$-spaces associated to a vector measure. We also study the Weyl transform\nof vector-valued functions and prove a vector-valued analogue of the\nHausdorff-Young inequality.\n", " The \"Up-and-down\" theorem which describes the structure of the Boolean\nalgebra of fragments of a linear positive operator is the well known result of\nthe operator theory. We prove an analog of this theorem for a positive abstract\nUryson operator defined on a vector lattice and taking values in a Dedekind\ncomplete vector lattice. This result we apply to prove that for an order narrow\npositive abstract Uryson operator $T$ from a vector lattice $E$ to a Dedekind\ncomplete vector lattice $F$, every abstract Uryson operator $S:E\\to F$, such\nthat $0\\leq S\\leq T$ is also order narrow.\n", " For the free Fermion gas at thermodynamic equilibrium, the temperature which\nrepresents the averaged kinetic energy is demonstrated to have the proper\nproperty which is invariant under the Lorentz boost, using the conceptions of\nthe relativistic quantum field theory and statistical physics. The\nthermodynamic interaction is described by the coupling of particles with the\nvacuum of a scalar pseudo boson which features a self-interaction. The various\nmeasurement of the apparent temperature of a moving body is also investigated.\n", " We propose a method that morphs high-orger meshes such that their boundaries\nand interfaces coincide/align with implicitly defined geometries. Our focus is\nparticularly on the case when the target surface is prescribed as the zero\nisocontour of a smooth discrete function. Common examples of this scenario\ninclude using level set functions to represent material interfaces in\nmultimaterial configurations, and evolving geometries in shape and topology\noptimization. The proposed method formulates the mesh optimization problem as a\nvariational minimization of the sum of a chosen mesh-quality metric using the\nTarget-Matrix Optimization Paradigm (TMOP) and a penalty term that weakly\nforces the selected faces of the mesh to align with the target surface. The\ndistinct features of the method are use of a source mesh to represent the level\nset function with sufficient accuracy, and adaptive strategies for setting the\npenalization weight and selecting the faces of the mesh to be fit to the target\nisocontour of the level set field. We demonstrate that the proposed method is\nrobust for generating boundary- and interface-fitted meshes for curvilinear\ndomains using different element types in 2D and 3D.\n", " In this paper, we prove a theorem of linearized asymptotic stability for\nfractional differential equations with a time delay. More precisely, using the\nmethod of linearization of a nonlinear equation along an orbit (Lyapunov's\nfirst method), we show that an equilibrium of a nonlinear Caputo fractional\ndifferential equation with a time delay is asymptotically stable if its\nlinearization at the equilibrium is asymptotically stable. Our approach based\non a technique which converts the linear part of the equation into a diagonal\none. Then using properties of generalized Mittag-Leffler functions, the\nconstruction of an associated Lyapunov--Perron operator and the Banach\ncontraction mapping theorem, we obtain the desired result.\n", " We find that studying the simplest of the coupled non-equilibrium growth\nequations of Barabasi by self-consistent mode coupling requires the use of\ndressed vertices. Using the vertex renormalization, we find a roughness\nexponent which already in the leading order is quite close to the numerical\nvalue.\n", " The Lie algebra generated by Hopf-zero classical normal forms is decomposed\ninto two versal Lie subalgebras. Some dynamical properties for each subalgebra\nare described; one is the set of all volume-preserving conservative systems\nwhile the other is the maximal Lie algebra of nonconservative systems. This\nintroduces a unique conservative--nonconservative decomposition for the normal\nform systems. There exists a Lie--subalgebra that is Lie-isomorphic to a large\nfamily of vector fields with Bogdanov--Takens singularity. This gives rise to a\nconclusion that the local dynamics of formal Hopf-zero singularities is\nwell-understood by the study of Bogdanov--Takens singularities. Despite this,\nthe normal form computation of Bogdanov-Takens and Hopf-zero singularities are\nindependent. Thus, by assuming a quadratic non-zero condition, complete results\non the simplest Hopf-zero normal forms are obtained in terms of the\nconservative--nonconservative decomposition. Some practical formulas are\nderived and the results implemented using Maple. The method has been applied on\nthe R\\\"ossler and Kuramoto--Sivashinsky equations to demonstrate the\napplicability of our results.\n" ]

[ " The many-body Green's function provides access to electronic properties\nbeyond density functional theory level in ab inito calculations. In this\nmanuscript, we propose a deep learning framework for predicting the\nfinite-temperature Green's function in atomic orbital space, aiming to achieve\na balance between accuracy and efficiency. By predicting the self-energy\nmatrices in Lehmann representation using an equivariant message passing neural\nnetwork, our method respects its analytical property and the $E(3)$\nequivariance. The Green's function is obtained from the predicted self-energy\nthrough Dyson equation with target total number of electrons. We present\nproof-of-concept benchmark results for both molecules and simple periodic\nsystems, showing that our method is able to provide accurate estimate of\nphysical observables such as energy and density of states based on the\npredicted Green's function.\n", " In this paper, we propose a new framework, exploiting the multi-agent deep\ndeterministic policy gradient (MADDPG) algorithm, to enable a base station (BS)\nand user equipment (UE) to come up with a medium access control (MAC) protocol\nin a multiple access scenario. In this framework, the BS and UEs are\nreinforcement learning (RL) agents that need to learn to cooperate in order to\ndeliver data. The network nodes can exchange control messages to collaborate\nand deliver data across the network, but without any prior agreement on the\nmeaning of the control messages. In such a framework, the agents have to learn\nnot only the channel access policy, but also the signaling policy. The\ncollaboration between agents is shown to be important, by comparing the\nproposed algorithm to ablated versions where either the communication between\nagents or the central critic is removed. The comparison with a contention-free\nbaseline shows that our framework achieves a superior performance in terms of\ngoodput and can effectively be used to learn a new protocol.\n", " Systems biology is an inter-disciplinary field that studies systems of\nbiological components at different scales, which may be molecules, cells or\nentire organism. In particular, systems biology methods are applied to\nunderstand functional deregulations within human cells (e.g., cancers). In this\ncontext, we present several python packages linked to CellNOptR (R package),\nwhich is used to build predictive logic models of signalling networks by\ntraining networks (derived from literature) to signalling (phospho-proteomic)\ndata. The first package (cellnopt.wrapper) is a wrapper based on RPY2 that\nallows a full access to CellNOptR functionalities within Python. The second one\n(cellnopt.core) was designed to ease the manipulation and visualisation of data\nstructures used in CellNOptR, which was achieved by using Pandas, NetworkX and\nmatplotlib. Systems biology also makes extensive use of web resources and\nservices. We will give an overview and status of BioServices, which allows one\nto access programmatically to web resources used in life science and how it can\nbe combined with CellNOptR.\n", " Humans spend over 90% of their time in buildings which account for 40% of\nanthropogenic greenhouse gas (GHG) emissions, making buildings the leading\ncause of climate change. To incentivize more sustainable construction, building\ncodes are used to enforce indoor comfort standards and maximum energy use.\nHowever, they currently only reward energy efficiency measures such as\nequipment or envelope upgrades and disregard the actual spatial configuration\nand usage. Using a new hypergraph model that encodes building floorplan\norganization and facilitates automatic geometry creation, we demonstrate that\nspace efficiency outperforms envelope upgrades in terms of operational carbon\nemissions in 72%, 61% and 33% of surveyed buildings in Zurich, New York, and\nSingapore. Automatically generated floorplans for a case study in Zurich\nfurther increase access to daylight by up to 24%, revealing that auto-generated\nfloorplans have the potential to improve the quality of residential spaces in\nterms of environmental performance and access to daylight.\n", " The DifSets package for GAP implements an algorithm for enumerating all\ndifference sets in a group up to equivalence and provides access to a library\nof results. The algorithm functions by finding difference sums, which are\npotential images of difference sets in quotient groups of the original group,\nand searching their preimages. In this way, the search space can be\ndramatically decreased, and searches of groups of relatively large order (such\nas order 64 or order 96) can be completed.\n", " Embedded non-volatile memory technologies such as resistive random access\nmemory (RRAM) and spin-transfer torque magnetic RAM (STT MRAM) are increasingly\nbeing researched for application in neuromorphic computing and hardware\naccelerators for AI. However, the stochastic write processes in these memory\ntechnologies affect their yield and need to be studied alongside process\nvariations, which drastically increase the complexity of yield analysis using\nthe Monte Carlo approach. Therefore, we propose an approach based on the\nFokker-Planck equation for modeling the stochastic write processes in STT MRAM\nand RRAM devices. Moreover, we show that our proposed approach can reproduce\nthe experimental results for both STT-MRAM and RRAM devices.\n", " To address growth challenges facing large Data Centers and supercomputing\nclusters a new construction is presented for scalable, high throughput, low\nlatency networks. The resulting networks require 1.5-5 times fewer switches,\n2-6 times fewer cables, have 1.2-2 times lower latency and correspondingly\nlower congestion and packet losses than the best present or proposed networks\nproviding the same number of ports at the same total bisection. These advantage\nratios increase with network size. The key new ingredient is the exact\nequivalence discovered between the problem of maximizing network bisection for\nlarge classes of practically interesting Cayley graphs and the problem of\nmaximizing codeword distance for linear error correcting codes. Resulting\ntranslation recipe converts existent optimal error correcting codes into\noptimal throughput networks.\n", " In the wake of the burgeoning expansion of generative artificial intelligence\n(AI) services, the computational demands inherent to these technologies\nfrequently necessitate cloud-powered computational offloading, particularly for\nresource-constrained mobile devices. These services commonly employ prompts to\nsteer the generative process, and both the prompts and the resultant content,\nsuch as text and images, may harbor privacy-sensitive or confidential\ninformation, thereby elevating security and privacy risks. To mitigate these\nconcerns, we introduce $\\Lambda$-Split, a split computing framework to\nfacilitate computational offloading while simultaneously fortifying data\nprivacy against risks such as eavesdropping and unauthorized access. In\n$\\Lambda$-Split, a generative model, usually a deep neural network (DNN), is\npartitioned into three sub-models and distributed across the user's local\ndevice and a cloud server: the input-side and output-side sub-models are\nallocated to the local, while the intermediate, computationally-intensive\nsub-model resides on the cloud server. This architecture ensures that only the\nhidden layer outputs are transmitted, thereby preventing the external\ntransmission of privacy-sensitive raw input and output data. Given the\nblack-box nature of DNNs, estimating the original input or output from\nintercepted hidden layer outputs poses a significant challenge for malicious\neavesdroppers. Moreover, $\\Lambda$-Split is orthogonal to traditional\nencryption-based security mechanisms, offering enhanced security when deployed\nin conjunction. We empirically validate the efficacy of the $\\Lambda$-Split\nframework using Llama 2 and Stable Diffusion XL, representative large language\nand diffusion models developed by Meta and Stability AI, respectively. Our\n$\\Lambda$-Split implementation is publicly accessible at\nhttps://github.com/nishio-laboratory/lambda_split.\n", " Federated Learning (FL) has emerged as a promising technique for edge devices\nto collaboratively learn a shared machine learning model while keeping training\ndata locally on the device, thereby removing the need to store and access the\nfull data in the cloud. However, FL is difficult to implement, test and deploy\nin practice considering heterogeneity in common edge device settings, making it\nfundamentally hard for researchers to efficiently prototype and test their\noptimization algorithms. In this work, our aim is to alleviate this problem by\nintroducing FL_PyTorch : a suite of open-source software written in python that\nbuilds on top of one the most popular research Deep Learning (DL) framework\nPyTorch. We built FL_PyTorch as a research simulator for FL to enable fast\ndevelopment, prototyping and experimenting with new and existing FL\noptimization algorithms. Our system supports abstractions that provide\nresearchers with a sufficient level of flexibility to experiment with existing\nand novel approaches to advance the state-of-the-art. Furthermore, FL_PyTorch\nis a simple to use console system, allows to run several clients simultaneously\nusing local CPUs or GPU(s), and even remote compute devices without the need\nfor any distributed implementation provided by the user. FL_PyTorch also offers\na Graphical User Interface. For new methods, researchers only provide the\ncentralized implementation of their algorithm. To showcase the possibilities\nand usefulness of our system, we experiment with several well-known\nstate-of-the-art FL algorithms and a few of the most common FL datasets.\n", " The utilization of web mapping becomes increasingly important in the domain\nof cartography. Users want access to spatial data on the web specific to their\nneeds. For this reason, different approaches were appeared for generating\non-the-fly the maps demanded by users, but those not suffice for guide a\nflexible and efficient process. Thus, new approach must be developed for\nimproving this process according to the user needs. This work focuses on\ndefining a new strategy which improves on-the-fly map generalization process\nand resolves the spatial conflicts. This approach uses the multiple\nrepresentation and cartographic generalization. The map generalization process\nis based on the implementation of multi- agent system where each agent was\nequipped with a genetic patrimony.\n", " Given a set S of n \\geq d points in general position in R^d, a random\nhyperplane split is obtained by sampling d points uniformly at random without\nreplacement from S and splitting based on their affine hull. A random\nhyperplane search tree is a binary space partition tree obtained by recursive\napplication of random hyperplane splits. We investigate the structural\ndistributions of such random trees with a particular focus on the growth with\nd. A blessing of dimensionality arises--as d increases, random hyperplane\nsplits more closely resemble perfectly balanced splits; in turn, random\nhyperplane search trees more closely resemble perfectly balanced binary search\ntrees.\n We prove that, for any fixed dimension d, a random hyperplane search tree\nstoring n points has height at most (1 + O(1/sqrt(d))) log_2 n and average\nelement depth at most (1 + O(1/d)) log_2 n with high probability as n\n\\rightarrow \\infty. Further, we show that these bounds are asymptotically\noptimal with respect to d.\n", " Deep Reinforcement Learning (RL) methods rely on experience replay to\napproximate the minibatched supervised learning setting; however, unlike\nsupervised learning where access to lots of training data is crucial to\ngeneralization, replay-based deep RL appears to struggle in the presence of\nextraneous data. Recent works have shown that the performance of Deep Q-Network\n(DQN) degrades when its replay memory becomes too large.\n This suggests that outdated experiences somehow impact the performance of\ndeep RL, which should not be the case for off-policy methods like DQN.\nConsequently, we re-examine the motivation for sampling uniformly over a replay\nmemory, and find that it may be flawed when using function approximation. We\nshow that -- despite conventional wisdom -- sampling from the uniform\ndistribution does not yield uncorrelated training samples and therefore biases\ngradients during training. Our theory prescribes a special non-uniform\ndistribution to cancel this effect, and we propose a stratified sampling scheme\nto efficiently implement it.\n", " Recent observations suggest the presence of clouds in exoplanet atmospheres\nbut have also shown that certain chemical species in the upper atmosphere might\nnot be in chemical equilibrium. The goal of this work is to calculate the two\nmain cloud formation processes, nucleation and bulk growth, consistently from a\nnon-equilibrium gas-phase. The aim is further to explore the interaction\nbetween a kinetic gas-phase and cloud micro-physics. The cloud formation is\nmodeled using the moment method and kinetic nucleation which are coupled to a\ngas-phase kinetic rate network. Specifically, the formation of cloud\ncondensation nuclei is derived from cluster rates that include the\nthermochemical data of (TiO$_2$)$_N$ from N = 1 to 15. The surface growth of 9\nbulk Al/Fe/Mg/O/Si/S/Ti binding materials considers the respective gas-phase\nspecies through condensation and surface reactions as derived from kinetic\ndisequilibrium. The effect of completeness of rate networks and the time\nevolution of the cloud particle formation is studied for an example exoplanet\nHD 209458 b. A consistent, fully time-dependent cloud formation model in\nchemical disequilibrium with respect to nucleation, bulk growth and the\ngas-phase is presented and first test cases are studied. This model shows that\ncloud formation in exoplanet atmospheres is a fast process. This confirms\nprevious findings that the formation of cloud particles is a local process.\nTests on selected locations within the atmosphere of the gas-giant HD 209458 b\nshow that the cloud particle number density and volume reach constant values\nwithin 1s. The complex kinetic polymer nucleation of TiO$_2$ confirms results\nfrom classical nucleation models. The surface reactions of SiO[s] and\nSiO$_2$[s] can create a catalytic cycle that dissociates H$_2$ to 2 H,\nresulting in a reduction of the CH$_4$ number densities.\n", " We introduce an orientation-preserving landmark-based distance for continuous\ncurves, which can be viewed as an alternative to the \\Frechet or Dynamic Time\nWarping distances. This measure retains many of the properties of those\nmeasures, and we prove some relations, but can be interpreted as a Euclidean\ndistance in a particular vector space. Hence it is significantly easier to use,\nfaster for general nearest neighbor queries, and allows easier access to\nclassification results than those measures. It is based on the \\emph{signed}\ndistance function to the curves or other objects from a fixed set of landmark\npoints. We also prove new stability properties with respect to the choice of\nlandmark points, and along the way introduce a concept called signed local\nfeature size (slfs) which parameterizes these notions. Slfs explains the\ncomplexity of shapes such as non-closed curves where the notion of local\norientation is in dispute -- but is more general than the well-known concept of\n(unsigned) local feature size, and is for instance infinite for closed simple\ncurves. Altogether, this work provides a novel, simple, and powerful method for\noriented shape similarity and analysis.\n", " The third-order response lies at the heart of simulating and interpreting\nnonlinear spectroscopies ranging from two dimensional infrared (2D-IR) to 2D\nelectronic (2D-ES), and 2D sum frequency generation (2D-SFG). The extra time\nand frequency dimensions in these spectroscopies provides access to rich\ninformation on the electronic and vibrational states present, the coupling\nbetween them, and the resulting rates at which they exchange energy that are\nobscured in linear spectroscopy, particularly for condensed phase systems that\nusually contain many overlapping features. While the exact quantum expression\nfor the third-order response is well established it is incompatible with the\nmethods that are practical for calculating the atomistic dynamics of large\ncondensed phase systems. These methods, which include both classical mechanics\nand quantum dynamics methods that retain quantum statistical properties while\nobeying the symmetries of classical dynamics, such as LSC-IVR, Centroid\nMolecular Dynamics (CMD) and Ring Polymer Molecular Dynamics (RPMD) naturally\nprovide short-time approximations to the multi-time symmetrized Kubo\ntransformed correlation function. Here, we show how the third-order response\ncan be formulated in terms of equilibrium symmetrized Kubo transformed\ncorrelation functions. We demonstrate the utility and accuracy of our approach\nby showing how it can be used to obtain the third-order response of a series of\nmodel systems using both classical dynamics and RPMD. In particular, we show\nthat this approach captures features such as anharmonically induced vertical\nsplittings and peak shifts while providing a physically transparent framework\nfor understanding multidimensional spectroscopies.\n", " If observations confirm BICEP2's claim of a tensor-scalar ratio $r\\approx\n0.2$ on CMB scales, then the inflationary consistency relation $n_{t}=-r/8$\npredicts a small negative value for the tensor spectral index $n_t$. We show\nthat future CMB polarization experiments should be able to confirm this\nprediction at several sigma. We also show how to properly extend the\nconsistency relation to solar system scales, where the primordial gravitational\nwave density $\\Omega_{gw}$ could be measured by proposed experiments such as\nthe Big Bang Observer. This would provide a far more stringent test of the\nconsistency relation and access much more detailed information about the early\nuniverse.\n", " Biomedical networks (or graphs) are universal descriptors for systems of\ninteracting elements, from molecular interactions and disease co-morbidity to\nhealthcare systems and scientific knowledge. Advances in artificial\nintelligence, specifically deep learning, have enabled us to model, analyze,\nand learn with such networked data. In this review, we put forward an\nobservation that long-standing principles of systems biology and medicine --\nwhile often unspoken in machine learning research -- provide the conceptual\ngrounding for representation learning on graphs, explain its current successes\nand limitations, and even inform future advancements. We synthesize a spectrum\nof algorithmic approaches that, at their core, leverage graph topology to embed\nnetworks into compact vector spaces. We also capture the breadth of ways in\nwhich representation learning has dramatically improved the state-of-the-art in\nbiomedical machine learning. Exemplary domains covered include identifying\nvariants underlying complex traits, disentangling behaviors of single cells and\ntheir effects on health, assisting in diagnosis and treatment of patients, and\ndeveloping safe and effective medicines.\n", " Within a fully microscopic setting, we derive a variational principle for the\nnon-equilibrium steady states of chemical reaction networks, valid for\ntime-scales over which chemical potentials can be taken to be slowly varying:\nat stationarity the system minimizes a global function of the reaction fluxes\nwith the form of a Hopfield Hamiltonian with Hebbian couplings, that is\nexplicitly seen to correspond to the rate of decay of entropy production over\ntime. Guided by this analogy, we show that reaction networks can be formally\nre-cast as systems of interacting reactions that optimize the use of the\navailable compounds by competing for substrates, akin to agents competing for a\nlimited resource in an optimal allocation problem. As an illustration, we\nanalyze the scenario that emerges in two simple cases: that of toy (random)\nreaction networks and that of a metabolic network model of the human red blood\ncell.\n", " In the paper we apply graph vertex coloring for verification of secret\nshares. We start from showing how to convert any graph into the number and vice\nversa. Next, theoretical result concerning properties of n-colorable graphs is\nstated and proven. From this result we derive graph coloring check-digit\nscheme. Feasibility of proposed scheme increases with the size of the number,\nwhich digits are checked and overall probability of errors. The check-digit\nscheme is used to build shares verification method that does not require\ncooperation of the third party. It allows implementing verification structure\ndifferent from the access structure. It does not depend on particular secret\nsharing method. It can be used as long as the secret shares can be represented\nby numbers or graphs.\n", " Often noisy point clouds are given as an approximation of a particular\ncompact set of interest. A finite point cloud is a compact set. This paper\nproves a reconstruction theorem which gives a sufficient condition, as a bound\non the Hausdorff distance between two compact sets, for when certain offsets of\nthese two sets are homotopic in terms of the absence of {\\mu}-critical points\nin an annular region. Since an offset of a set deformation retracts to the set\nitself provided that there are no critical points of the distance function\nnearby, we can use this theorem to show when the offset of a point cloud is\nhomotopy equivalent to the set it is sampled from. The ambient space can be any\nRiemannian manifold but we focus on ambient manifolds which have nowhere\nnegative curvature. In the process, we prove stability theorems for\n{\\mu}-critical points when the ambient space is a manifold.\n", " The present study focuses on the interaction of gravity waves in the\natmosphere with the tropopause. As the vertical extent of the latter is small\ncompared to the density scale height, wave propagation is described by the\nTaylor-Goldstein equation as derived from the linearised Boussinesq\napproximation. Of particular interest in the construction of gravity wave\nparameterisations for the upper atmosphere are the transmission and reflection\nproperties of the tropopause as these determine the upward fluxes of energy and\nmomentum carried by internal waves.\n A method is presented that decomposes internal waves explicitly into upward\nand downward propagating contributions, thus giving direct access to\ntransmission and reflection coefficients of finite regions of non-uniform\nstratification in a stationary atmosphere. The scheme utilizes a piecewise\nconstant approximation for the background stratification and matches up- and\ndownward propagating plane wave solutions in each layer through physically\nmeaningful coupling conditions. As a result, transmission and reflection\ncoefficients follow immediately.\n", " The complex interactions involved in regulation of a cell's function are\ncaptured by its interaction graph. More often than not, detailed knowledge\nabout enhancing or suppressive regulatory influences and cooperative effects is\nlacking and merely the presence or absence of directed interactions is known.\nHere we investigate to which extent such reduced information allows to forecast\nthe effect of a knock-out or a combination of knock-outs. Specifically we ask\nin how far the lethality of eliminating nodes may be predicted by their network\ncentrality, such as degree and betweenness, without knowing the function of the\nsystem. The function is taken as the ability to reproduce a fixed point under a\ndiscrete Boolean dynamics. We investigate two types of stochastically generated\nnetworks: fully random networks and structures grown with a mechanism of node\nduplication and subsequent divergence of interactions. On all networks we find\nthat the out-degree is a good predictor of the lethality of a single node\nknock-out. For knock-outs of node pairs, the fraction of successors shared\nbetween the two knocked-out nodes (out-overlap) is a good predictor of\nsynthetic lethality. Out-degree and out-overlap are locally defined and\ncomputationally simple centrality measures that provide a predictive power\nclose to the optimal predictor.\n", " Movable antenna (MA) is a promising technology to improve wireless\ncommunication performance by varying the antenna position in a given finite\narea at the transceivers to create more favorable channel conditions. In this\npaper, we investigate the MA-enhanced multiple-access channel (MAC) for the\nuplink transmission from multiple users each equipped with a single MA to a\nbase station (BS) with a fixed-position antenna (FPA) array. A field-response\nbased channel model is used to characterize the multi-path channel between the\nantenna array of the BS and each user's MA with a flexible position. To\nevaluate the MAC performance gain provided by MAs, we formulate an optimization\nproblem for minimizing the total transmit power of users, subject to a\nminimum-achievable-rate requirement for each user, where the positions of MAs\nand the transmit powers of users, as well as the receive combining matrix of\nthe BS are jointly optimized. To solve this non-convex optimization problem\ninvolving intricately coupled variables, we develop two algorithms based on\nzero-forcing (ZF) and minimum mean square error (MMSE) combining methods,\nrespectively. Specifically, for each algorithm, the combining matrix of the BS\nand the total transmit power of users are expressed as a function of the MAs'\nposition vectors, which are then optimized by using the proposed\nmulti-directional descent (MDD) framework. It is shown that the proposed\nZF-based and MMSE-based MDD algorithms can converge to high-quality suboptimal\nsolutions with low computational complexities. Simulation results demonstrate\nthat the proposed solutions for MA-enhanced multiple access systems can\nsignificantly decrease the total transmit power of users as compared to\nconventional FPA systems employing antenna selection under both perfect and\nimperfect field-response information.\n", " Data Science aims to extract meaningful knowledge from unorganised data. Real\ndatasets usually come in the form of a cloud of points with only pairwise\ndistances. Numerous applications require to visualise an overall shape of a\nnoisy cloud of points sampled from a non-linear object that is more complicated\nthan a union of disjoint clusters. The skeletonisation problem in its hardest\nform is to find a 1-dimensional skeleton that correctly represents a shape of\nthe cloud. This paper compares several algorithms that solve the above\nskeletonisation problem for any point cloud and guarantee a successful\nreconstruction. For example, given a highly noisy point sample of an unknown\nunderlying graph, a reconstructed skeleton should be geometrically close and\nhomotopy equivalent to (has the same number of independent cycles as) the\nunderlying graph. One of these algorithm produces a Homologically Persistent\nSkeleton (HoPeS) for any cloud without extra parameters. This universal\nskeleton contains sub-graphs that provably represent the 1-dimensional shape of\nthe cloud at any scale. Other subgraphs of HoPeS reconstruct an unknown graph\nfrom its noisy point sample with a correct homotopy type and within a small\noffset of the sample. The extensive experiments on synthetic and real data\nreveal for the first time the maximum level of noise that allows successful\ngraph reconstructions.\n", " With the rapid advancement of 5G networks, billions of smart Internet of\nThings (IoT) devices along with an enormous amount of data are generated at the\nnetwork edge. While still at an early age, it is expected that the evolving 6G\nnetwork will adopt advanced artificial intelligence (AI) technologies to\ncollect, transmit, and learn this valuable data for innovative applications and\nintelligent services. However, traditional machine learning (ML) approaches\nrequire centralizing the training data in the data center or cloud, raising\nserious user-privacy concerns. Federated learning, as an emerging distributed\nAI paradigm with privacy-preserving nature, is anticipated to be a key enabler\nfor achieving ubiquitous AI in 6G networks. However, there are several system\nand statistical heterogeneity challenges for effective and efficient FL\nimplementation in 6G networks. In this article, we investigate the optimization\napproaches that can effectively address the challenging heterogeneity issues\nfrom three aspects: incentive mechanism design, network resource management,\nand personalized model optimization. We also present some open problems and\npromising directions for future research.\n", " Signaling in enzymatic networks is typically triggered by environmental\nfluctuations, resulting in a series of stochastic chemical reactions, leading\nto corruption of the signal by noise. For example, information flow is\ninitiated by binding of extracellular ligands to receptors, which is\ntransmitted through a {cascade involving} kinase-phosphatase stochastic\nchemical reactions. For a class of such networks, we develop a general\nfield-theoretic approach in order to calculate the error in signal transmission\nas a function of an appropriate control variable. Application of the theory to\na simple push-pull network, a module in the kinase-phosphatase cascade,\nrecovers the exact results for error in signal transmission previously obtained\nusing umbral calculus (Phys. Rev. X., {\\bf 4}, 041017 (2014)). We illustrate\nthe generality of the theory by studying the minimal errors in noise reduction\nin a reaction cascade with two connected push-pull modules. Such a cascade\nbehaves as an effective {three-species} network with a pseudo intermediate. In\nthis case, optimal information transfer, resulting in the smallest square of\nthe error between the input and output, occurs with a time delay, which is\ngiven by the inverse of the decay rate of the pseudo intermediate. There are\nsubstantial deviations between simulations and predictions of the linear theory\nin error in signal propagation in an enzymatic push-pull network for a certain\nrange of parameters. Inclusion of second order perturbative corrections shows\nthat differences between simulations and theoretical predictions are minimized.\nOur study establishes that {a} field theoretic formulation {of} stochastic\nbiological {signaling offers} a systematic way to understand error propagation\nin networks of arbitrary complexity.\n", " Modern Neural Network (NN) architectures heavily rely on vast numbers of\nmultiply-accumulate arithmetic operations, constituting the predominant\ncomputational cost. Therefore, this paper proposes a high-throughput, scalable\nand energy efficient non-element-wise matrix multiplication unit on FPGAs as a\nbasic component of the NNs. We firstly streamline inter-layer and intra-layer\nredundancies of MADDNESS algorithm, a LUT-based approximate matrix\nmultiplication, to design a fast, efficient scalable approximate matrix\nmultiplication module termed \"Approximate Multiplication Unit (AMU)\". The AMU\noptimizes LUT-based matrix multiplications further through dedicated memory\nmanagement and access design, decoupling computational overhead from input\nresolution and boosting FPGA-based NN accelerator efficiency significantly. The\nexperimental results show that using our AMU achieves up to 9x higher\nthroughput and 112x higher energy efficiency over the state-of-the-art\nsolutions for the FPGA-based Quantised Neural Network (QNN) accelerators.\n", " The tree reconstruction problem is to find an embedded straight-line tree\nthat approximates a given cloud of unorganized points in $\\mathbb{R}^m$ up to a\ncertain error. A practical solution to this problem will accelerate a discovery\nof new colloidal products with desired physical properties such as viscosity.\nWe define the Approximate Skeleton of any finite point cloud $C$ in a Euclidean\nspace with theoretical guarantees. The Approximate Skeleton ASk$(C)$ always\nbelongs to a given offset of $C$, i.e. the maximum distance from $C$ to\nASk$(C)$ can be a given maximum error. The number of vertices in the\nApproximate Skeleton is close to the minimum number in an optimal tree by\nfactor 2. The new Approximate Skeleton of any unorganized point cloud $C$ is\ncomputed in a near linear time in the number of points in $C$. Finally, the\nApproximate Skeleton outperforms past skeletonization algorithms on the size\nand accuracy of reconstruction for a large dataset of real micelles and random\nclouds.\n", " With ever increasing data rates produced by modern radio telescopes like\nLOFAR and future telescopes like the SKA, many data processing steps are\noverwhelmed by the amount of data that needs to be handled using limited\ncompute resources. Calibration is one such operation that dominates the overall\ndata processing computational cost, nonetheless, it is an essential operation\nto reach many science goals. Calibration algorithms do exist that scale well\nwith the number of stations of an array and the number of directions being\ncalibrated. However, the remaining bottleneck is the raw data volume, which\nscales with the number of baselines, and which is proportional to the square of\nthe number of stations. We propose a 'stochastic' calibration strategy where we\nonly read in a mini-batch of data for obtaining calibration solutions, as\nopposed to reading the full batch of data being calibrated. Nonetheless, we\nobtain solutions that are valid for the full batch of data. Normally, data need\nto be averaged before calibration is performed to accommodate the data in\nsize-limited compute memory. Stochastic calibration overcomes the need for data\naveraging before any calibration can be performed, and offers many advantages\nincluding: enabling the mitigation of faint radio frequency interference;\nbetter removal of strong celestial sources from the data; and better detection\nand spatial localization of fast radio transients.\n", " This paper studies the mathematical properties of collectively canalizing\nBoolean functions, a class of functions that has arisen from applications in\nsystems biology. Boolean networks are an increasingly popular modeling\nframework for regulatory networks, and the class of functions studied here\ncaptures a key feature of biological network dynamics, namely that a subset of\none or more variables, under certain conditions, can dominate the value of a\nBoolean function, to the exclusion of all others. These functions have rich\nmathematical properties to be explored. The paper shows how the number and type\nof such sets influence a function's behavior and define a new measure for the\ncanalizing strength of any Boolean function. We further connect the concept of\ncollective canalization with the well-studied concept of the average\nsensitivity of a Boolean function. The relationship between Boolean functions\nand the dynamics of the networks they form is important in a wide range of\napplications beyond biology, such as computer science, and has been studied\nwith statistical and simulation-based methods. But the rich relationship\nbetween structure and dynamics remains largely unexplored, and this paper is\nintended as a contribution to its mathematical foundation.\n", " e-Yantra Robotics Competition (eYRC) is a unique Robotics Competition hosted\nby IIT Bombay that is actually an Embedded Systems and Robotics MOOC.\nRegistrations have been growing exponentially in each year from 4500 in 2012 to\nover 34000 in 2019. In this 5-month long competition students learn complex\nskills under severe time pressure and have access to a discussion forum to post\ndoubts about the learning material. Responding to questions in real-time is a\nchallenge for project staff. Here, we illustrate the advantage of Deep Learning\nfor real-time question answering in the eYRC discussion forum. We illustrate\nthe advantage of Transformer based contextual embedding mechanisms such as\nBidirectional Encoder Representation From Transformer (BERT) over word\nembedding mechanisms such as Word2Vec. We propose a weighted similarity metric\nas a measure of matching and find it more reliable than Content-Content or\nTitle-Title similarities alone. The automation of replying to questions has\nbrought the turn around response time(TART) down from a minimum of 21 mins to a\nminimum of 0.3 secs.\n", " This paper is concerned with the utilization of deterministically modeled\nchemical reaction networks for the implementation of (feed-forward) neural\nnetworks. We develop a general mathematical framework and prove that the\nordinary differential equations (ODEs) associated with certain reaction network\nimplementations of neural networks have desirable properties including (i)\nexistence of unique positive fixed points that are smooth in the parameters of\nthe model (necessary for gradient descent), and (ii) fast convergence to the\nfixed point regardless of initial condition (necessary for efficient\nimplementation). We do so by first making a connection between neural networks\nand fixed points for systems of ODEs, and then by constructing reaction\nnetworks with the correct associated set of ODEs. We demonstrate the theory by\nconstructing a reaction network that implements a neural network with a\nsmoothed ReLU activation function, though we also demonstrate how to generalize\nthe construction to allow for other activation functions (each with the\ndesirable properties listed previously). As there are multiple types of\n\"networks\" utilized in this paper, we also give a careful introduction to both\nreaction networks and neural networks, in order to disambiguate the overlapping\nvocabulary in the two settings and to clearly highlight the role of each\nnetwork's properties.\n", " Massive datasets and high-capacity models have driven many recent\nadvancements in computer vision and natural language understanding. This work\npresents a platform to enable similar success stories in Embodied AI. We\npropose ProcTHOR, a framework for procedural generation of Embodied AI\nenvironments. ProcTHOR enables us to sample arbitrarily large datasets of\ndiverse, interactive, customizable, and performant virtual environments to\ntrain and evaluate embodied agents across navigation, interaction, and\nmanipulation tasks. We demonstrate the power and potential of ProcTHOR via a\nsample of 10,000 generated houses and a simple neural model. Models trained\nusing only RGB images on ProcTHOR, with no explicit mapping and no human task\nsupervision produce state-of-the-art results across 6 embodied AI benchmarks\nfor navigation, rearrangement, and arm manipulation, including the presently\nrunning Habitat 2022, AI2-THOR Rearrangement 2022, and RoboTHOR challenges. We\nalso demonstrate strong 0-shot results on these benchmarks, via pre-training on\nProcTHOR with no fine-tuning on the downstream benchmark, often beating\nprevious state-of-the-art systems that access the downstream training data.\n", " Living cells can enhance their fitness by anticipating environmental change.\nWe study how accurately linear signaling networks in cells can predict future\nsignals. We find that maximal predictive power results from a combination of\ninput-noise suppression, linear extrapolation, and selective readout of\ncorrelated past signal values. Single-layer networks generate exponential\nresponse kernels, which suffice to predict Markovian signals optimally.\nMultilayer networks allow oscillatory kernels that can optimally predict\nnon-Markovian signals. At low noise, these kernels exploit the signal\nderivative for extrapolation, while at high noise, they capitalize on signal\nvalues in the past that are strongly correlated with the future signal. We show\nhow the common motifs of negative feedback and incoherent feed-forward can\nimplement these optimal response functions. Simulations reveal that E. coli can\nreliably predict concentration changes for chemotaxis, and that the integration\ntime of its response kernel arises from a trade-off between rapid response and\nnoise suppression.\n", " Neural network training is usually accomplished by solving a non-convex\noptimization problem using stochastic gradient descent. Although one optimizes\nover the networks parameters, the main loss function generally only depends on\nthe realization of the neural network, i.e. the function it computes. Studying\nthe optimization problem over the space of realizations opens up new ways to\nunderstand neural network training. In particular, usual loss functions like\nmean squared error and categorical cross entropy are convex on spaces of neural\nnetwork realizations, which themselves are non-convex. Approximation\ncapabilities of neural networks can be used to deal with the latter\nnon-convexity, which allows us to establish that for sufficiently large\nnetworks local minima of a regularized optimization problem on the realization\nspace are almost optimal. Note, however, that each realization has many\ndifferent, possibly degenerate, parametrizations. In particular, a local\nminimum in the parametrization space needs not correspond to a local minimum in\nthe realization space. To establish such a connection, inverse stability of the\nrealization map is required, meaning that proximity of realizations must imply\nproximity of corresponding parametrizations. We present pathologies which\nprevent inverse stability in general, and, for shallow networks, proceed to\nestablish a restricted space of parametrizations on which we have inverse\nstability w.r.t. to a Sobolev norm. Furthermore, we show that by optimizing\nover such restricted sets, it is still possible to learn any function which can\nbe learned by optimization over unrestricted sets.\n", " Living cells can leverage correlations in environmental fluctuations to\npredict the future environment and mount a response ahead of time. To this end,\ncells need to encode the past signal into the output of the intracellular\nnetwork from which the future input is predicted. Yet, storing information is\ncostly while not all features of the past signal are equally informative on the\nfuture input signal. Here, we show, for two classes of input signals, that\ncellular networks can reach the fundamental bound on the predictive information\nas set by the information extracted from the past signal: push-pull networks\ncan reach this information bound for Markovian signals, while networks that\ntake a temporal derivative can reach the bound for predicting the future\nderivative of non-Markovian signals. However, the bits of past information that\nare most informative about the future signal are also prohibitively costly. As\na result, the optimal system that maximizes the predictive information for a\ngiven resource cost is, in general, not at the information bound. Applying our\ntheory to the chemotaxis network of Escherichia coli reveals that its adaptive\nkernel is optimal for predicting future concentration changes over a broad\nrange of background concentrations, and that the system has been tailored to\npredicting these changes in shallow gradients.\n", " Network Design problems typically ask for a minimum cost sub-network from a\ngiven host network. This classical point-of-view assumes a central authority\nenforcing the optimum solution. But how should networks be designed to cope\nwith selfish agents that own parts of the network? In this setting, agents will\ndeviate from a minimum cost network if this decreases their individual cost.\nHence, designed networks should be both efficient in terms of total cost and\nstable in terms of the agents' willingness to accept the network.\n We study this novel type of Network Design problem by investigating the\ncreation of $(\\beta,\\gamma)$-networks, that are in $\\beta$-approximate Nash\nequilibrium and have a total cost of at most $\\gamma$ times the optimal cost,\nfor the recently proposed Euclidean Generalized Network Creation Game by Bil\\`o\net al. [SPAA 2019]. There, $n$ agents corresponding to points in Euclidean\nspace create costly edges among themselves to optimize their centrality in the\ncreated network. Our main result is a simple $\\mathcal{O}(n^2)$-time algorithm\nthat computes a $(\\beta,\\beta)$-network with low $\\beta$ for any given set of\npoints. Moreover, on integer grid point sets or random point sets our algorithm\nachieves a low constant $\\beta$. Besides these results, we discuss a\ngeneralization of our algorithm to instances with arbitrary, even non-metric,\nedge lengths. Moreover, we show that no such positive results are possible when\nfocusing on either optimal networks or perfectly stable networks as in both\ncases NP-hard problems arise, there exist instances with very unstable optimal\nnetworks, and there are instances for perfectly stable networks with high total\ncost. Along the way, we significantly improve several results from Bil\\`o et\nal. and we asymptotically resolve their conjecture about the Price of Anarchy\nby providing a tight bound.\n", " This paper presents a comparative study of six soft computing models namely\nmultilayer perceptron networks, Elman recurrent neural network, radial basis\nfunction network, Hopfield model, fuzzy inference system and hybrid fuzzy\nneural network for the hourly electricity demand forecast of Czech Republic.\nThe soft computing models were trained and tested using the actual hourly load\ndata for seven years. A comparison of the proposed techniques is presented for\npredicting 2 day ahead demands for electricity. Simulation results indicate\nthat hybrid fuzzy neural network and radial basis function networks are the\nbest candidates for the analysis and forecasting of electricity demand.\n", " The susceptibility of deep learning to adversarial attack can be understood\nin the framework of the Renormalisation Group (RG) and the vulnerability of a\nspecific network may be diagnosed provided the weights in each layer are known.\nAn adversary with access to the inputs and outputs could train a second network\nto clone these weights and, having identified a weakness, use them to compute\nthe perturbation of the input data which exploits it. However, the RG framework\nalso provides a means to poison the outputs of the network imperceptibly,\nwithout affecting their legitimate use, so as to prevent such cloning of its\nweights and thereby foil the generation of adversarial data.\n", " Analog beamforming is an attractive and cost-effective solution to exploit\nthe benefits of massive multiple-input-multiple-output systems, by requiring\nonly one up/down-conversion chain. However, the presence of only one chain\nimposes a significant overhead in estimating the channel state information\nrequired for beamforming, when conventional digital channel estimation (CE)\napproaches are used. As an alternative, this paper proposes a novel CE\ntechnique, called periodic analog CE (PACE), that can be performed by analog\nhardware. By avoiding digital processing, the estimation overhead is\nsignificantly lowered and does not scale with number of antennas. PACE involves\nperiodic transmission of a sinusoidal reference signal by the transmitter,\nestimation of its amplitude and phase at each receive antenna via analog\nhardware, and using these estimates for beamforming. To enable such non-trivial\noperation, two reference tone recovery techniques and a novel receiver\narchitecture for PACE are proposed and analyzed, both theoretically and via\nsimulations. Results suggest that in sparse, wide-band channels and above a\ncertain signal-to-noise ratio, PACE aided beamforming suffers only a small loss\nin beamforming gain and enjoys a much lower CE overhead, in comparison to\nconventional approaches. Benefits of using PACE aided beamforming during the\ninitial access phase are also discussed.\n", " The recent discovery of methods to isolate graphene, a one-atom-thick layer\nof crystalline carbon, has raised the possibility of a new class of\nnano-electronics devices based on the extraordinary electrical transport and\nunusual physical properties of this material. However, the experimental\nrealization of devices displaying these properties was, until now, hampered by\nthe influence of the ambient environment, primarily the substrate. Here we\nreport on the fabrication of Suspended Graphene devices and on studies of their\nelectrical transport properties. In these devices, environmental disturbances\nwere minimized allowing unprecedented access to the intrinsic properties of\ngraphene close to the Dirac Point (DP) where the energy dispersion of the\ncarriers and their density-of-states vanish linearly giving rise to a range of\nexotic physical properties. We show that charge inhomogeneity is reduced by\nalmost one order of magnitude compared to that in Non-Suspended Graphene\ndevices. Moreover, near the DP, the mobility exceeds 100,000 cm2/Vs,\napproaching theoretical predictions for evanescent transport in the ballistic\nmodel.\n", " We derive fundamental lower bounds on the connectivity and the memory\nrequirements of deep neural networks guaranteeing uniform approximation rates\nfor arbitrary function classes in $L^2(\\mathbb R^d)$. In other words, we\nestablish a connection between the complexity of a function class and the\ncomplexity of deep neural networks approximating functions from this class to\nwithin a prescribed accuracy. Additionally, we prove that our lower bounds are\nachievable for a broad family of function classes. Specifically, all function\nclasses that are optimally approximated by a general class of representation\nsystems---so-called \\emph{affine systems}---can be approximated by deep neural\nnetworks with minimal connectivity and memory requirements. Affine systems\nencompass a wealth of representation systems from applied harmonic analysis\nsuch as wavelets, ridgelets, curvelets, shearlets, $\\alpha$-shearlets, and more\ngenerally $\\alpha$-molecules. Our central result elucidates a remarkable\nuniversality property of neural networks and shows that they achieve the\noptimum approximation properties of all affine systems combined. As a specific\nexample, we consider the class of $\\alpha^{-1}$-cartoon-like functions, which\nis approximated optimally by $\\alpha$-shearlets. We also explain how our\nresults can be extended to the case of functions on low-dimensional immersed\nmanifolds. Finally, we present numerical experiments demonstrating that the\nstandard stochastic gradient descent algorithm generates deep neural networks\nproviding close-to-optimal approximation rates. Moreover, these results\nindicate that stochastic gradient descent can actually learn approximations\nthat are sparse in the representation systems optimally sparsifying the\nfunction class the network is trained on.\n", " Convolutional Neural Networks (CNN) have been successful in processing data\nsignals that are uniformly sampled in the spatial domain (e.g., images).\nHowever, most data signals do not natively exist on a grid, and in the process\nof being sampled onto a uniform physical grid suffer significant aliasing error\nand information loss. Moreover, signals can exist in different topological\nstructures as, for example, points, lines, surfaces and volumes. It has been\nchallenging to analyze signals with mixed topologies (for example, point cloud\nwith surface mesh). To this end, we develop mathematical formulations for\nNon-Uniform Fourier Transforms (NUFT) to directly, and optimally, sample\nnonuniform data signals of different topologies defined on a simplex mesh into\nthe spectral domain with no spatial sampling error. The spectral transform is\nperformed in the Euclidean space, which removes the translation ambiguity from\nworks on the graph spectrum. Our representation has four distinct advantages:\n(1) the process causes no spatial sampling error during the initial sampling,\n(2) the generality of this approach provides a unified framework for using CNNs\nto analyze signals of mixed topologies, (3) it allows us to leverage\nstate-of-the-art backbone CNN architectures for effective learning without\nhaving to design a particular architecture for a particular data structure in\nan ad-hoc fashion, and (4) the representation allows weighted meshes where each\nelement has a different weight (i.e., texture) indicating local properties. We\nachieve results on par with the state-of-the-art for the 3D shape retrieval\ntask, and a new state-of-the-art for the point cloud to surface reconstruction\ntask.\n", " Ethanediol is one of the largest complex organic molecules detected in space\nthus far. It has been found in different types of molecular clouds. The two\npropanediol isomers are the next larger diols. Hence, they are viable\ncandidates to be searched for in space. We wish to provide sufficiently large\nand accurate sets of spectroscopic parameters of 1,2-propanediol to facilitate\nsearches for this molecule at millimeter and longer submillimeter wavelengths.\nWe recorded rotational spectra of 1,2-propanediol in three wide frequency\nwindows between 38 and 400~GHz. We made extensive assignments for the three\nlowest energy conformers to yield spectroscopic parameters up to eighth order\nof angular momentum. Our present data will be helpful for identifying\n1,2-propanediol at moderate submillimeter or longer wavelengths with radio\ntelescope arrays such as ALMA, NOEMA, or EVLA. In particular, its detection\nwith ALMA in sources, in which ethanediol was detected, appears to be\npromising.\n", " The similarity transformed equation of motion coupled cluster (STEOM-CC)\nmethod is benchmarked against CC3 and EOM-CCSDT-3 for a large test set of\nvalence excited states of organic molecules studied by Schreiber et al. [M.\nSchreiber, M.R. Silva-Junior, S.P. Sauer, and W. Thiel, J. Chem. Phys.\n$\\textbf{128}$, 134110 (2008)]. STEOM-CC is found to behave quite\nsatisfactorily and provides significant improvement over EOM-CCSD, CASPT2 and\nNEVPT2 for singlet excited states, lowering standard deviations of these\nmethods by almost a factor of 2. Triplet excited states are found to be\ndescribed less accurately, however. Besides the parent version of STEOM-CC,\nadditional variations are considered. STEOM-D includes a perturbative\ncorrection from doubly excited determinants. The novel STEOM-H ($\\omega$)\napproach presents a sophisticated technique to render the STEOM-CC transformed\nHamiltonian hermitian. In STEOM-PT, the expensive CCSD step is replaced by\nmany-body second-order perturbation theory (MBPT(2)), while extended STEOM\n(EXT-STEOM) provides access to doubly excited states. To study orbital\ninvariance in STEOM, we investigate orbital rotation in the STEOM-ORB approach.\nComparison of theses variations of STEOM for the large test set provides a\ncomprehensive statistical basis to gauge the usefulness of these approaches.\n", " In wireless networks, energy consumed for communication includes both the\ntransmission and the processing energy. In this paper, point-to-point\ncommunication over a fading channel with an energy harvesting transmitter is\nstudied considering jointly the energy costs of transmission and processing.\nUnder the assumption of known energy arrival and fading profiles, optimal\ntransmission policy for throughput maximization is investigated. Assuming that\nthe transmitter has sufficient amount of data in its buffer at the beginning of\nthe transmission period, the average throughput by a given deadline is\nmaximized. Furthermore, a \"directional glue pouring algorithm\" that computes\nthe optimal transmission policy is described.\n", " In this work the collapsing process of a spherically symmetric star, made of\ndust cloud, in the background of dark energy is studied for two different\ngravity theories separately, i.e., DGP Brane gravity and Loop Quantum gravity.\nTwo types of dark energy fluids, namely, Modified Chaplygin gas and Generalised\nCosmic Chaplygin gas are considered for each model. Graphs are drawn to\ncharacterize the nature and the probable outcome of gravitational collapse. A\ncomparative study is done between the collapsing process in the two different\ngravity theories. It is found that in case of dark matter, there is a great\npossibility of collapse and consequent formation of Black hole. In case of dark\nenergy possibility of collapse is far lesser compared to the other cases, due\nto the large negative pressure of dark energy component. There is an increase\nin mass of the cloud in case of dark matter collapse due to matter\naccumulation. The mass decreases considerably in case of dark energy due to\ndark energy accretion on the cloud. In case of collapse with a combination of\ndark energy and dark matter, it is found that in the absence of interaction\nthere is a far better possibility of formation of black hole in DGP brane model\ncompared to Loop quantum cosmology model.\n", " One of the first widespread uses of multi-user multiple-input multiple-output\n(MU-MIMO) is in 5G networks, where each base station has an advanced antenna\nsystem (AAS) that is connected to the baseband unit (BBU) with a\ncapacity-constrained fronthaul. In the AAS configuration, multiple passive\nantenna elements and radio units are integrated into a single box. This paper\nconsiders precoded downlink transmission over a single-cell MU-MIMO system. We\nstudy optimized linear precoding for AAS with a limited-capacity fronthaul,\nwhich requires the precoding matrix to be quantized. We propose a new precoding\ndesign that is aware of the fronthaul quantization and minimizes the\nmean-squared error at the receiver side. We compute the precoding matrix using\na sphere decoding (SD) approach. We also propose a heuristic low-complexity\napproach to quantized precoding. This heuristic is computationally efficient\nenough for massive MIMO systems. The numerical results show that our proposed\nprecoding significantly outperforms quantization-unaware precoding and other\nprevious approaches in terms of the sum rate. The performance loss for our\nheuristic method compared to quantization-aware precoding is insignificant\nconsidering the complexity reduction, which makes the heuristic method feasible\nfor real-time applications. We consider both perfect and imperfect channel\nstate information.\n", " I postulate that human or other intelligent agents function or should\nfunction as follows. They store all sensory observations as they come - the\ndata is holy. At any time, given some agent's current coding capabilities, part\nof the data is compressible by a short and hopefully fast program / description\n/ explanation / world model. In the agent's subjective eyes, such data is more\nregular and more \"beautiful\" than other data. It is well-known that knowledge\nof regularity and repeatability may improve the agent's ability to plan actions\nleading to external rewards. In absence of such rewards, however, known beauty\nis boring. Then \"interestingness\" becomes the first derivative of subjective\nbeauty: as the learning agent improves its compression algorithm, formerly\napparently random data parts become subjectively more regular and beautiful.\nSuch progress in compressibility is measured and maximized by the curiosity\ndrive: create action sequences that extend the observation history and yield\npreviously unknown / unpredictable but quickly learnable algorithmic\nregularity. We discuss how all of the above can be naturally implemented on\ncomputers, through an extension of passive unsupervised learning to the case of\nactive data selection: we reward a general reinforcement learner (with access\nto the adaptive compressor) for actions that improve the subjective\ncompressibility of the growing data. An unusually large breakthrough in\ncompressibility deserves the name \"discovery\". The \"creativity\" of artists,\ndancers, musicians, pure mathematicians can be viewed as a by-product of this\nprinciple. Several qualitative examples support this hypothesis.\n", " The ribosome flow model with input and output (RFMIO) is a deterministic\ndynamical system that has been used to study the flow of ribosomes during mRNA\ntranslation. The RFMIO and its variants encapsulate important properties that\nare relevant to modeling ribosome flow such as the possible evolution of\n\"traffic jams\" and non-homogeneous elongation rates along the mRNA molecule,\nand can also be used for studying additional intracellular processes such as\ntranscription, transport, and more. Here we consider networks of interconnected\nRFMIOs as a fundamental tool for modeling, analyzing and re-engineering the\ncomplex mechanisms of protein production. In these networks, the output of each\nRFMIO may be divided, using connection weights, between several inputs of other\nRFMIOs. We show that under quite general feedback connections the network has\ntwo important properties: (1) it admits a unique steady-state and every\ntrajectory converges to this steady-state, and (2) the problem of how to\ndetermine the connection weights so that the network steady-state output is\nmaximized is a convex optimization problem. These mathematical properties make\nthese networks highly suitable as models of various phenomena: property (1)\nmeans that the behavior is predictable and ordered, and property (2) means that\ndetermining the optimal weights is numerically tractable even for large-scale\nnetworks. For the specific case of a feed-forward network of RFMIOs we prove an\nadditional useful property, namely, that there exists a spectral representation\nfor the network steady-state, and thus it can be determined without any\nnumerical simulations of the dynamics. We describe the implications of these\nresult to several fundamental biological phenomena and biotechnological\nobjectives.\n" ]

[ " In this paper we analyze the performance of 2*4 MIMO-MC-CDMA system in MATLAB\nwhich highly reduces BER. In this paper we combine MIMO and MC-CDMA system to\nreduce bit error rate in which MC-CDMA is multi user and multiple access\nschemes which is used to increase the data rate of the system. MC-CDMA system\nis a single wideband frequency selective carrier which converts frequency\nselective to parallel narrowband flat fading multiple sub-carriers to enhance\nthe performance of system. Now MC-CDMA system further improved by grouping with\n2*4 MIMO system which uses ZF (Zero Forcing) decoder at the receiver to\ndecrease BER with half rate convolutionally encoded Alamouti STBC block code is\nused as transmit diversity of MIMO through multiple transmit antenna.\nImportance of using MIMO-MC-CDMA using convolution code is firstly to reduce\nthe complexity of system secondary to reduce BER and lastly to increase gain.\nIn this paper we examine system performance in diverse modulation techniques\nlike, 8-PSK, 16-QAM, QPSK, 32-QAM, 8-QAM and 64-QAM in Rayleigh fading channel\nusing MATLAB.\n", " One of the first widespread uses of multi-user multiple-input multiple-output\n(MU-MIMO) is in 5G networks, where each base station has an advanced antenna\nsystem (AAS) that is connected to the baseband unit (BBU) with a\ncapacity-constrained fronthaul. In the AAS configuration, multiple passive\nantenna elements and radio units are integrated into a single box. This paper\nconsiders precoded downlink transmission over a single-cell MU-MIMO system. We\nstudy optimized linear precoding for AAS with a limited-capacity fronthaul,\nwhich requires the precoding matrix to be quantized. We propose a new precoding\ndesign that is aware of the fronthaul quantization and minimizes the\nmean-squared error at the receiver side. We compute the precoding matrix using\na sphere decoding (SD) approach. We also propose a heuristic low-complexity\napproach to quantized precoding. This heuristic is computationally efficient\nenough for massive MIMO systems. The numerical results show that our proposed\nprecoding significantly outperforms quantization-unaware precoding and other\nprevious approaches in terms of the sum rate. The performance loss for our\nheuristic method compared to quantization-aware precoding is insignificant\nconsidering the complexity reduction, which makes the heuristic method feasible\nfor real-time applications. We consider both perfect and imperfect channel\nstate information.\n", " An accurate and efficient forecasting system is imperative to the prevention\nof emerging infectious diseases such as COVID-19 in public health. This system\nrequires accurate transient modeling, lower computation cost, and fewer\nobservation data. To tackle these three challenges, we propose a novel deep\nlearning approach using black-box knowledge distillation for both accurate and\nefficient transmission dynamics prediction in a practical manner. First, we\nleverage mixture models to develop an accurate, comprehensive, yet impractical\nsimulation system. Next, we use simulated observation sequences to query the\nsimulation system to retrieve simulated projection sequences as knowledge.\nThen, with the obtained query data, sequence mixup is proposed to improve query\nefficiency, increase knowledge diversity, and boost distillation model\naccuracy. Finally, we train a student deep neural network with the retrieved\nand mixed observation-projection sequences for practical use. The case study on\nCOVID-19 justifies that our approach accurately projects infections with much\nlower computation cost when observation data are limited.\n", " It is now well established that the cell is a highly crowded environment.\nYet, the effects of crowding on the dynamics of signaling pathways, gene\nregulation networks and metabolic networks are still largely unknown. Crowding\ncan alter both molecular diffusion and the equilibria of biomolecular\nreactions. In this review, we first discuss how diffusion can affect\nbiochemical networks. Diffusion of transcription factors can increase noise in\ngene expression, while diffusion of proteins between intracellular compartments\nor between cells can reduce concentration fluctuations. In push-pull networks\ndiffusion can impede information transmission, while in multi-site protein\nmodification networks diffusion can qualitatively change the macroscopic\nresponse of the system, such as the loss or emergence of bistability. Moreover,\ndiffusion can directly change the metabolic flux. We describe how crowding\naffects diffusion, and thus how all these phenomena are influenced by crowding.\nYet, a potentially more important effect of crowding on biochemical networks is\nmediated via the shift in the equilibria of bimolecular reactions, and we\nprovide computational evidence that supports this idea. Finally, we discuss how\nthe effects of crowding can be incorporated in models of biochemical networks.\n", " In wireless networks, energy consumed for communication includes both the\ntransmission and the processing energy. In this paper, point-to-point\ncommunication over a fading channel with an energy harvesting transmitter is\nstudied considering jointly the energy costs of transmission and processing.\nUnder the assumption of known energy arrival and fading profiles, optimal\ntransmission policy for throughput maximization is investigated. Assuming that\nthe transmitter has sufficient amount of data in its buffer at the beginning of\nthe transmission period, the average throughput by a given deadline is\nmaximized. Furthermore, a \"directional glue pouring algorithm\" that computes\nthe optimal transmission policy is described.\n", " In the simplest view of transcriptional regulation, the expression of a gene\nis turned on or off by changes in the concentration of a transcription factor\n(TF). We use recent data on noise levels in gene expression to show that it\nshould be possible to transmit much more than just one regulatory bit.\nRealizing this optimal information capacity would require that the dynamic\nrange of TF concentrations used by the cell, the input/output relation of the\nregulatory module, and the noise levels of binding and transcription satisfy\ncertain matching relations. This parameter-free prediction is in good agreement\nwith recent experiments on the Bicoid/Hunchback system in the early Drosophila\nembryo, and this system achieves ~90% of its theoretical maximum information\ntransmission.\n", " In this paper we analyse the construction of identification codes.\nIdentification codes are based on the question \"Is the message I have just\nreceived the one I am interested in?\", as opposed to Shannon's transmission,\nwhere the receiver is interested in not only one, but any message. The\nadvantage of identification is that it allows rates growing double\nexponentially in the blocklength at the cost of not being able to decode every\nmessage, which might be beneficial in certain applications. We focus on a\nspecial identification code construction based on two concatenated Reed-Solomon\ncodes and have a closer look at its implementation, analyzing the trade-offs of\nidentification with respect to transmission and the trade-offs introduced by\nthe computational cost of identification codes.\n", " One of the most important parts of Artificial Neural Networks is minimizing\nthe loss functions which tells us how good or bad our model is. To minimize\nthese losses we need to tune the weights and biases. Also to calculate the\nminimum value of a function we need gradient. And to update our weights we need\ngradient descent. But there are some problems with regular gradient descent ie.\nit is quite slow and not that accurate. This article aims to give an\nintroduction to optimization strategies to gradient descent. In addition, we\nshall also discuss the architecture of these algorithms and further\noptimization of Neural Networks in general\n", " Strong correlation can be essentially captured with multireference\nwavefunction methods such as complete active space self-consistent field\n(CASSCF) or density matrix renormalization group (DMRG). Still, an accurate\ndescription of the electronic structure of strongly correlated systems requires\naccounting for the dynamic electron correlation, which CASSCF and DMRG largely\nmiss. In this work a new approach for the correlation energy based on the\nadiabatic connection (AC) is proposed. The AC$_{\\rm n}$ method accounts for\nterms up to the desired order n in the coupling constant, is rigorously\nsize-consistent, free from instabilities and intruder states. It employs the\nparticle-hole multireference random phase approximation and the Cholesky\ndecomposition technique, which leads to a computational cost growing with the\nfifth power of the system size. Thanks to AC$_{\\rm n}$ depending solely on one-\nand two-electron CAS reduced density matrix, the method is much more efficient\nthan existing ab initio dynamic correlation methods for strong correlation.\nAC$_{\\rm n}$ affords excellent results for singlet-triplet gaps of challenging\norganic biradicals. Development presented in this work opens new perspectives\nfor accurate calculations of systems with dozens of strongly correlated\nelectrons.\n", " Movable antenna (MA) is a promising technology to improve wireless\ncommunication performance by varying the antenna position in a given finite\narea at the transceivers to create more favorable channel conditions. In this\npaper, we investigate the MA-enhanced multiple-access channel (MAC) for the\nuplink transmission from multiple users each equipped with a single MA to a\nbase station (BS) with a fixed-position antenna (FPA) array. A field-response\nbased channel model is used to characterize the multi-path channel between the\nantenna array of the BS and each user's MA with a flexible position. To\nevaluate the MAC performance gain provided by MAs, we formulate an optimization\nproblem for minimizing the total transmit power of users, subject to a\nminimum-achievable-rate requirement for each user, where the positions of MAs\nand the transmit powers of users, as well as the receive combining matrix of\nthe BS are jointly optimized. To solve this non-convex optimization problem\ninvolving intricately coupled variables, we develop two algorithms based on\nzero-forcing (ZF) and minimum mean square error (MMSE) combining methods,\nrespectively. Specifically, for each algorithm, the combining matrix of the BS\nand the total transmit power of users are expressed as a function of the MAs'\nposition vectors, which are then optimized by using the proposed\nmulti-directional descent (MDD) framework. It is shown that the proposed\nZF-based and MMSE-based MDD algorithms can converge to high-quality suboptimal\nsolutions with low computational complexities. Simulation results demonstrate\nthat the proposed solutions for MA-enhanced multiple access systems can\nsignificantly decrease the total transmit power of users as compared to\nconventional FPA systems employing antenna selection under both perfect and\nimperfect field-response information.\n", " This paper is concerned with the utilization of deterministically modeled\nchemical reaction networks for the implementation of (feed-forward) neural\nnetworks. We develop a general mathematical framework and prove that the\nordinary differential equations (ODEs) associated with certain reaction network\nimplementations of neural networks have desirable properties including (i)\nexistence of unique positive fixed points that are smooth in the parameters of\nthe model (necessary for gradient descent), and (ii) fast convergence to the\nfixed point regardless of initial condition (necessary for efficient\nimplementation). We do so by first making a connection between neural networks\nand fixed points for systems of ODEs, and then by constructing reaction\nnetworks with the correct associated set of ODEs. We demonstrate the theory by\nconstructing a reaction network that implements a neural network with a\nsmoothed ReLU activation function, though we also demonstrate how to generalize\nthe construction to allow for other activation functions (each with the\ndesirable properties listed previously). As there are multiple types of\n\"networks\" utilized in this paper, we also give a careful introduction to both\nreaction networks and neural networks, in order to disambiguate the overlapping\nvocabulary in the two settings and to clearly highlight the role of each\nnetwork's properties.\n", " With the rapid advancement of 5G networks, billions of smart Internet of\nThings (IoT) devices along with an enormous amount of data are generated at the\nnetwork edge. While still at an early age, it is expected that the evolving 6G\nnetwork will adopt advanced artificial intelligence (AI) technologies to\ncollect, transmit, and learn this valuable data for innovative applications and\nintelligent services. However, traditional machine learning (ML) approaches\nrequire centralizing the training data in the data center or cloud, raising\nserious user-privacy concerns. Federated learning, as an emerging distributed\nAI paradigm with privacy-preserving nature, is anticipated to be a key enabler\nfor achieving ubiquitous AI in 6G networks. However, there are several system\nand statistical heterogeneity challenges for effective and efficient FL\nimplementation in 6G networks. In this article, we investigate the optimization\napproaches that can effectively address the challenging heterogeneity issues\nfrom three aspects: incentive mechanism design, network resource management,\nand personalized model optimization. We also present some open problems and\npromising directions for future research.\n", " The third-order response lies at the heart of simulating and interpreting\nnonlinear spectroscopies ranging from two dimensional infrared (2D-IR) to 2D\nelectronic (2D-ES), and 2D sum frequency generation (2D-SFG). The extra time\nand frequency dimensions in these spectroscopies provides access to rich\ninformation on the electronic and vibrational states present, the coupling\nbetween them, and the resulting rates at which they exchange energy that are\nobscured in linear spectroscopy, particularly for condensed phase systems that\nusually contain many overlapping features. While the exact quantum expression\nfor the third-order response is well established it is incompatible with the\nmethods that are practical for calculating the atomistic dynamics of large\ncondensed phase systems. These methods, which include both classical mechanics\nand quantum dynamics methods that retain quantum statistical properties while\nobeying the symmetries of classical dynamics, such as LSC-IVR, Centroid\nMolecular Dynamics (CMD) and Ring Polymer Molecular Dynamics (RPMD) naturally\nprovide short-time approximations to the multi-time symmetrized Kubo\ntransformed correlation function. Here, we show how the third-order response\ncan be formulated in terms of equilibrium symmetrized Kubo transformed\ncorrelation functions. We demonstrate the utility and accuracy of our approach\nby showing how it can be used to obtain the third-order response of a series of\nmodel systems using both classical dynamics and RPMD. In particular, we show\nthat this approach captures features such as anharmonically induced vertical\nsplittings and peak shifts while providing a physically transparent framework\nfor understanding multidimensional spectroscopies.\n", " Motivation: Usefulness of analysis derived from Affymetrix microarrays\ndepends largely upon the reliability of files describing the correspondence\nbetween probe sets, genes and transcripts. In particular, in case a gene is\ntargeted by two probe sets, one must be able to assess if the corresponding\nsignals measure a group of common transcripts or two groups of transcripts with\nlittle or no overlap.\n Results: Probe sets that effectively target the same group of transcripts\nhave specific properties in the trancriptional networks we constructed. We\nfound indeed that such probe sets had a very low negative correlation, a high\npositive correlation and a similar neighbourhood. Taking advantage of these\nproperties, we devised a test allowing to group probe sets which target the\nsame group of transcripts in a particular network. By considering several\nnetworks, additional information concerning the frequency of these associations\nwas obtained.\n Availability and Implementation: The programs developed in Python (PSAWNpy)\nand in Matlab (PSAWNml) are freely available, and can be downloaded at\nhttp://code.google.com/p/arraymatic/.\n Tutorials and reference manuals are available at\nhttp://bns.crbm.cnrs.fr/softwares.html.\n Contact: [email protected].\n Supplementary information: Supplementary data are available at\nhttp://bns.crbm.cnrs.fr/download.html.\n", " We study the following two problems:\n (1) Given $n\\ge 2$ and $\\al$, how large Hausdorff dimension can a compact set\n$A\\su\\Rn$ have if $A$ does not contain three points that form an angle $\\al$?\n (2) Given $\\al$ and $\\de$, how large Hausdorff dimension can a %compact\nsubset $A$ of a Euclidean space have if $A$ does not contain three points that\nform an angle in the $\\de$-neighborhood of $\\al$?\n An interesting phenomenon is that different angles show different behaviour\nin the above problems. Apart from the clearly special extreme angles 0 and\n$180^\\circ$, the angles $60^\\circ,90^\\circ$ and $120^\\circ$ also play special\nrole in problem (2): the maximal dimension is smaller for these special angles\nthan for the other angles. In problem (1) the angle $90^\\circ$ seems to behave\ndifferently from other angles.\n", " Phylogenetic diversity indices are commonly used to rank the elements in a\ncollection of species or populations for conservation purposes. The derivation\nof these indices is typically based on some quantitative description of the\nevolutionary history of the species in question, which is often given in terms\nof a phylogenetic tree. Both rooted and unrooted phylogenetic trees can be\nemployed, and there are close connections between the indices that are derived\nin these two different ways. In this paper, we introduce more general\nphylogenetic diversity indices that can be derived from collections of subsets\n(clusters) and collections of bipartitions (splits) of the given set of\nspecies. Such indices could be useful, for example, in case there is some\nuncertainty in the topology of the tree being used to derive a phylogenetic\ndiversity index. As well as characterizing some of the indices that we\nintroduce in terms of their special properties, we provide a link between\ncluster-based and split-based phylogenetic diversity indices that uses a\ndiscrete analogue of the classical link between affine and projective geometry.\nThis provides a unified framework for many of the various phylogenetic\ndiversity indices used in the literature based on rooted and unrooted\nphylogenetic trees, generalizations and new proofs for previous results\nconcerning tree-based indices, and a way to define some new phylogenetic\ndiversity indices that naturally arise as affine or projective variants of each\nother.\n", " Signal transduction pathways recover a crucial role in cellular processes:\nthey represent a connection between environmental conditions and cellular\nreactions. There are many pathways and they all are related to create a\nnetwork. But how can every protein find the right way more fast than possible?\nHow can it find the right down-streaming kinase in the cellular sea and not\nanother very similar kinase? Every signal transduction pathway can be seen as\ntwo distincted processes: the signal must reach every kinase and then it must\ntravel through the enzyme until its active site: quantum walks could be the\nanswer to both the questions.\n", " Nowadays, there is pressing demand for sustainable energy sources, or clean\nand 'green' fuel and hydrogen is a perfect candidate. It can be made by\ndissociating methane with the energy input compensated by metal-hydrogen bond\nformation. Industry uses Nickel supported on $\\gamma$-alumina for this\nrate-limiting step.\n This work shows how to use Quantum Monte Carlo calculations to take\n(user-guided) information summarising chemistry of heterogeneous catalysis from\na generic Jastrow factor, that allows electron correlation to be included in\nthe density. This, in turn, gives accurate values of properties, in particular\nthe activation barrier for stretching one methane C-H bond, before adsorbing\nCH3- and H- on vicinal nickel atoms in the close-packed (111) plane.\n The CH3- co-ordinates to a hollow site defined by three Ni atoms and, new to\nthis work, another Ni-atom co-ordinates the H-atom from the dissociation.\n", " Define an optimizer as having memory $k$ if it stores $k$ dynamically\nchanging vectors in the parameter space. Classical SGD has memory $0$, momentum\nSGD optimizer has $1$ and Adam optimizer has $2$. We address the following\nquestions: How can optimizers make use of more memory units? What information\nshould be stored in them? How to use them for the learning steps? As an\napproach to the last question, we introduce a general method called\n\"Retrospective Learning Law Correction\" or shortly RLLC. This method is\ndesigned to calculate a dynamically varying linear combination (called learning\nlaw) of memory units, which themselves may evolve arbitrarily. We demonstrate\nRLLC on optimizers whose memory units have linear update rules and small memory\n($\\leq 4$ memory units). Our experiments show that in a variety of standard\nproblems, these optimizers outperform the above mentioned three classical\noptimizers. We conclude that RLLC is a promising framework for boosting the\nperformance of known optimizers by adding more memory units and by making them\nmore adaptive.\n", " Protein phosphorylation is a reversible post-translational modification\ncommonly used by cell signaling networks to transmit information about the\nextracellular environment into intracellular organelles for the regulation of\nthe activity and sorting of proteins within the cell. For this study we\nreconstructed a literature-based mammalian kinase-substrate network from\nseveral online resources. The interactions within this directed graph network\nconnect kinases to their substrates, through specific phosphosites including\nkinase-kinase regulatory interactions. However, the \"signs\" of links,\nactivation or inhibition of the substrate upon phosphorylation, within this\nnetwork are mostly unknown. Here we show how we can infer the \"signs\"\nindirectly using data from quantitative phosphoproteomics experiments applied\nto mammalian cells combined with the literature-based kinase-substrate network.\nOur inference method was able to predict the sign for 321 links and 153\nphosphosites on 120 kinases, resulting in signed and directed subnetwork of\nmammalian kinase-kinase interactions. Such an approach can rapidly advance the\nreconstruction of cell signaling pathways and networks regulating mammalian\ncells.\n", " We characterize the measurement complexity of compressed sensing of signals\ndrawn from a known prior distribution, even when the support of the prior is\nthe entire space (rather than, say, sparse vectors). We show for Gaussian\nmeasurements and \\emph{any} prior distribution on the signal, that the\nposterior sampling estimator achieves near-optimal recovery guarantees.\nMoreover, this result is robust to model mismatch, as long as the distribution\nestimate (e.g., from an invertible generative model) is close to the true\ndistribution in Wasserstein distance. We implement the posterior sampling\nestimator for deep generative priors using Langevin dynamics, and empirically\nfind that it produces accurate estimates with more diversity than MAP.\n", " Signaling in enzymatic networks is typically triggered by environmental\nfluctuations, resulting in a series of stochastic chemical reactions, leading\nto corruption of the signal by noise. For example, information flow is\ninitiated by binding of extracellular ligands to receptors, which is\ntransmitted through a {cascade involving} kinase-phosphatase stochastic\nchemical reactions. For a class of such networks, we develop a general\nfield-theoretic approach in order to calculate the error in signal transmission\nas a function of an appropriate control variable. Application of the theory to\na simple push-pull network, a module in the kinase-phosphatase cascade,\nrecovers the exact results for error in signal transmission previously obtained\nusing umbral calculus (Phys. Rev. X., {\\bf 4}, 041017 (2014)). We illustrate\nthe generality of the theory by studying the minimal errors in noise reduction\nin a reaction cascade with two connected push-pull modules. Such a cascade\nbehaves as an effective {three-species} network with a pseudo intermediate. In\nthis case, optimal information transfer, resulting in the smallest square of\nthe error between the input and output, occurs with a time delay, which is\ngiven by the inverse of the decay rate of the pseudo intermediate. There are\nsubstantial deviations between simulations and predictions of the linear theory\nin error in signal propagation in an enzymatic push-pull network for a certain\nrange of parameters. Inclusion of second order perturbative corrections shows\nthat differences between simulations and theoretical predictions are minimized.\nOur study establishes that {a} field theoretic formulation {of} stochastic\nbiological {signaling offers} a systematic way to understand error propagation\nin networks of arbitrary complexity.\n", " Network analysis is currently used in a myriad of contexts: from identifying\npotential drug targets to predicting the spread of epidemics and designing\nvaccination strategies, and from finding friends to uncovering criminal\nactivity. Despite the promise of the network approach, the reliability of\nnetwork data is a source of great concern in all fields where complex networks\nare studied. Here, we present a general mathematical and computational\nframework to deal with the problem of data reliability in complex networks. In\nparticular, we are able to reliably identify both missing and spurious\ninteractions in noisy network observations. Remarkably, our approach also\nenables us to obtain, from those noisy observations, network reconstructions\nthat yield estimates of the true network properties that are more accurate than\nthose provided by the observations themselves. Our approach has the potential\nto guide experiments, to better characterize network data sets, and to drive\nnew discoveries.\n", " Data-driven dynamic models of cell biology can be used to predict cell\nresponse to unseen perturbations. Recent work (CellBox) had demonstrated the\nderivation of interpretable models with explicit interaction terms, in which\nthe parameters were optimized using machine learning techniques. While the\nprevious work was tested only in a single biological setting, this work aims to\nextend the range of applicability of this model inference approach to a\ndiversity of biological systems. Here we adapted CellBox in Julia differential\nprogramming and augmented the method with adjoint algorithms, which has\nrecently been used in the context of neural ODEs. We trained the models using\nsimulated data from both abstract and biology-inspired networks, which afford\nthe ability to evaluate the recovery of the ground truth network structure. The\nresulting accuracy of prediction by these models is high both in terms of low\nerror against data and excellent agreement with the network structure used for\nthe simulated training data. While there is no analogous ground truth for real\nlife biological systems, this work demonstrates the ability to construct and\nparameterize a considerable diversity of network models with high predictive\nability. The expectation is that this kind of procedure can be used on real\nperturbation-response data to derive models applicable to diverse biological\nsystems.\n", " We present the brief review of some central problems, anomalies, paradoxes of\nAstrophysics, Cosmology and new approaches to its decisions, proposed by the\nWave Universe Concept (WU Concept). All these, anyhow, - Hot Points of modern\nscience of Universe, about which Standard Model have not authentic answers. The\nrepresented set of brief (justified) Novels reflects very perspective\ndirections of search. Essential from thems - Headline (and Topics): New\nPhenomenon-Megaquantization: Observed Megaquantum Effects in Astronimical\nSystems. Internal Structure of Celestial Bodies and Sun. Where is disposes the\nConvective Zone? Towards to Stars: Where the Heliopause will be found? Mystery\nof the Fine Structure Constant (FSC). Answer of the WU Concept: Theoretical\nRepresentation of the Fine Structure Constant. FSC - As Micro and Mega\nParameter of the Universe. FSC, Orbits, Heliopause. What Quasars with Record\nRedshifts will be discovered in Future? Observed Universe: Monotonic\nHomogeneity and Anisotropy or Principal Hierarchy? Is the (Large - Scale) Limit\nof Universe exist? All answers, proposed by the Wave Universe Concept, in\ncontrast with speculative schemes of Standard Model, are effectively verified\nby experience, including, - and by justified in observations prognoses. It will\nbe waited, that namely here it is possible the real break in the understanding\nof these principal problems and enigmas, which Nature, infinite Universe offers\n(to us).\n", " In most multiple-input multiple-output (MIMO) communication systems, e.g.,\nMassive MIMO, the antenna spacing is generally no less than half a wavelength.\nIt helps to reduce the mutual coupling and therefore facilitate the system\ndesign. The maximum array gain is the number of antennas in this settings.\nHowever, when the antenna spacing is made very small, the array gain of a\ncompact array can be proportional to the square of the number of antennas - a\nvalue much larger than the traditional array. To achieve this so-called\n\"superdirectivity\" however, the calculation of the excitation coefficients\n(beamforming vector) is known to be a challenging problem. In this paper, we\nderive the beamforming vector of superdirective arrays based on a novel\ncoupling matrix-enabled method. We also propose an approach to obtain the\ncoupling matrix, which is derived by the spherical wave expansion method and\nactive element pattern. The full-wave electromagnetic simulations are conducted\nto validate the effectiveness of our proposed method. Simulation results show\nthat when the beamforming vector obtained by our method is applied, the\ndirectivity of the designed dipole antenna array has a good agreement with the\ntheoretical values.\n", " Background: A metabolic genotype comprises all chemical reactions an organism\ncan catalyze via enzymes encoded in its genome. A genotype is viable in a given\nenvironment if it is capable of producing all biomass components the organism\nneeds to survive and reproduce. Previous work has focused on the properties of\nindividual genotypes while little is known about how genome-scale metabolic\nnetworks with a given function can vary in their reaction content. Results: We\nhere characterize spaces of such genotypes. Specifically, we study metabolic\ngenotypes whose phenotype is viability in minimal chemical environments that\ndiffer in their sole carbon sources. We show that regardless of the number of\nreactions in a metabolic genotype, the genotypes of a given phenotype typically\nform vast, connected, and unstructured sets -- genotype networks -- that nearly\nspan the whole of genotype space. The robustness of metabolic phenotypes to\nrandom reaction removal in such spaces has a narrow distribution with a high\nmean. Different carbon sources differ in the number of metabolic genotypes in\ntheir genotype network; this number decreases as a genotype is required to be\nviable on increasing numbers of carbon sources, but much less than if metabolic\nreactions were used independently across different chemical environments.\nConclusions: Our work shows that phenotype-preserving genotype networks have\ngeneric organizational properties and that these properties are insensitive to\nthe number of reactions in metabolic genotypes.\n", " Without a doubt many problems in physics arise as a consequence of our\nphilosophical conception of the world. In this contribution however we endeavor\nto alleviate this scenario by putting forward a philosophical approach under\nwhich some of the most fundamental problems in modern physics might turn out to\nbe fictitious. To accomplish such task we propound that everything that exists\nmust be made up of \\emph{matter} which not only makes up space and the universe\nbut also is in constant change. For such reason the existence of total\nemptiness and material discontinuity are rejected. Here physical fields are\nassumed as a particular state of matter. And time is understood as the result\nof the intrinsic dynamics of the universe. Furthermore, the infiniteness of the\nuniverse is also discussed and its implications are briefly mentioned, e.g.,\nthe laws of conservation. Finally, the regularity of the physical laws is\nquestioned. In summary four great problems (from the perspective of physics)\nare suggested to be deeply studied: (1) What is matter?, (2) Why does the\nuniverse change? (3) is the universe infinite in extension? and (4) are there\nreally regular (invariant) laws of physics?\n", " Classification of targets by radar has proved to be notoriously difficult\nwith the best systems still yet to attain sufficiently high levels of\nperformance and reliability. In the current contribution we explore a new\ndesign of radar based target recognition, where angular diversity is used in a\ncognitive manner to attain better performance. Performance is bench- marked\nagainst conventional classification schemes. The proposed scheme can easily be\nextended to cognitive target recognition based on multiple diversity\nstrategies.\n", " Gene regulatory networks typically have low in-degrees, whereby any given\ngene is regulated by few of the genes in the network. What mechanisms might be\nresponsible for these low in-degrees? Starting with an accepted framework of\nthe binding of transcription factors to DNA, we consider a simple model of gene\nregulatory dynamics. In this model, we show that the constraint of having a\ngiven function leads to the emergence of minimum connectivities compatible with\nfunction. We exhibit mathematically this behavior within a limit of our model\nand show that it also arises in the full model. As a consequence, functionality\nin these gene networks is parsimonious, i.e., is concentrated on a sparse\nnumber of interactions as measured for instance by their essentiality. Our\nmodel thus provides a simple mechanism for the emergence of sparse regulatory\nnetworks, and leads to very heterogeneous effects of mutations.\n", " We study the asymptotic scaling properties of standard domain wall networks\nin several cosmological epochs. We carry out the largest field theory\nsimulations achieved to date, with simulation boxes of size 20483, and confirm\nthat a scale-invariant evolution of the network is indeed the attractor\nsolution. The simulations are also used to obtain an accurate calibration for\nthe velocity-dependent one-scale model for domain walls: we numerically\ndetermine the two free model parameters to have the values $c_w = 0.34\\pm0.16$\nand $k_w = 0.98\\pm0.07$, which are higher precision than (but in agreement\nwith) earlier estimates.\n", " In this paper, we consider the following graph embedding problem: Given a\nbipartite graph G = (V1; V2;E), where the maximum degree of vertices in V2 is\n4, can G be embedded on a two dimensional grid such that each vertex in V1 is\ndrawn as a line segment along a grid line, each vertex in V2 is drawn as a\npoint at a grid point, and each edge e = (u; v) for some u 2 V1 and v 2 V2 is\ndrawn as a line segment connecting u and v, perpendicular to the line segment\nfor u? We show that this problem is NP-complete, and sketch how our proof\ntechniques can be used to show the hardness of several other related problems.\n", " Gene expression is a noisy process that leads to regime shift between\nalternative steady states among individual living cells, inducing phenotypic\nvariability. The effects of white noise on the regime shift in bistable systems\nhave been well characterized, however little is known about such effects of\ncolored noise (noise with non-zero correlation time). Here, we show that noise\ncorrelation time, by considering a genetic circuit of autoactivation, can have\nsignificant effect on the regime shift in gene expression. We demonstrate this\ntheoretically, using stochastic potential, stationary probability density\nfunction and first-passage time based on the Fokker-Planck description, where\nthe Ornstein-Uhlenbeck process is used to model colored noise. We find that\nincrease in noise correlation time in degradation rate can induce a regime\nshift from low to high protein concentration state and enhance the bistable\nregime, while increase in noise correlation time in basal rate retain the\nbimodal distribution. We then show how cross-correlated colored noises in basal\nand degradation rates can induce regime shifts from low to high protein\nconcentration state, but reduce the bistable regime. In addition, we show that\nearly warning indicators can also be used to predict shifts between distinct\nphenotypic states in gene expression. Predictions that a cell is about to shift\nto a harmful phenotype could improve early therapeutic intervention in complex\nhuman diseases.\n", " One of the ultimate goals in biology is to understand the design principles\nof biological systems. Such principles, if they exist, can help us better\nunderstand complex, natural biological systems and guide the engineering of de\nnovo ones. Towards deciphering design principles, in silico evolution of\nbiological systems with proper abstraction is a promising approach. Here, we\ndemonstrate the application of in silico evolution combined with rule-based\nmodelling for exploring design principles of cellular signaling networks. This\napplication is based on a computational platform, called BioJazz, which allows\nin silico evolution of signaling networks with unbounded complexity. We provide\na detailed introduction to BioJazz architecture and implementation and describe\nhow it can be used to evolve and/or design signaling networks with defined\ndynamics. For the latter, we evolve signaling networks with switch-like\nresponse dynamics and demonstrate how BioJazz can result in new biological\ninsights on network structures that can endow bistable response dynamics. This\nexample also demonstrated both the power of BioJazz in evolving and designing\nsignaling networks and its limitations at the current stage of development.\n", " Devadoss asked: (1) can every polygon be convexified so that no internal\nvisibility (between vertices) is lost in the process? Moreover, (2) does such a\nconvexification exist, in which exactly one vertex is moved at a time (that is,\nusing {\\em single-vertex moves})? We prove the redundancy of the \"single-vertex\nmoves\" condition: an affirmative answer to (1) implies an affirmative answer to\n(2). Since Aichholzer et al. recently proved (1), this settles (2).\n", " Cellular processes do not follow deterministic rules; even in identical\nenvironments genetically identical cells can make random choices leading to\ndifferent phenotypes. This randomness originates from fluctuations present in\nthe biomolecular interaction networks. Most previous work has been focused on\nthe intrinsic noise (IN) of these networks. Yet, especially for\nhigh-copy-number biomolecules, extrinsic or environmental noise (EN) has been\nexperimentally shown to dominate the variation. Here, we develop an analytical\nformalism that allows for calculation of the effect of EN on gene-expression\nmotifs. We introduce a method for modeling bounded EN as an auxiliary species\nin the master equation. The method is fully generic and is not limited to\nsystems with small EN magnitudes. We focus our study on motifs that can be\nviewed as the building blocks of genetic switches: a nonregulated gene, a\nself-inhibiting gene, and a self-promoting gene. The role of the EN properties\n(magnitude, correlation time, and distribution) on the statistics of interest\nare systematically investigated, and the effect of fluctuations in different\nreaction rates is compared. Due to its analytical nature, our formalism can be\nused to quantify the effect of EN on the dynamics of biochemical networks and\ncan also be used to improve the interpretation of data from single-cell\ngene-expression experiments.\n", " Despite their topological complexity almost all functional properties of\nmetabolic networks can be derived from steady-state dynamics. Indeed, many\ntheoretical investigations (like flux-balance analysis) rely on extracting\nfunction from steady states. This leads to the interesting question, how\nmetabolic networks avoid complex dynamics and maintain a steady-state behavior.\nHere, we expose metabolic network topologies to binary dynamics generated by\nsimple local rules. We find that the networks' response is highly specific:\nComplex dynamics are systematically reduced on metabolic networks compared to\nrandomized networks with identical degree sequences. Already small topological\nmodifications substantially enhance the capacity of a network to host complex\ndynamic behavior and thus reduce its regularizing potential. This exceptionally\npronounced regularization of dynamics encoded in the topology may explain, why\nsteady-state behavior is ubiquitous in metabolism.\n", " A small number of (perhaps only 6) broken-symmetries, marked by the edges of\na hierarchical series of physical {\\em subsystem-types}, underlie the delicate\ncorrelation-based complexity of life on our planet's surface. Order-parameters\nassociated with these broken symmetries might in the future help us broaden our\ndefinitions of community health. For instance we show that a model of metazoan\nattention-focus, on correlation-layers that look in/out from the 3 boundaries\nof skin, family & culture, predicts that behaviorally-diverse communities\nrequire a characteristic task layer-multiplicity {\\em per individual} of only\nabout $4 \\frac14$ of the six correlation layers that comprise that community.\nThe model may facilitate explorations of task-layer diversity, go beyond GDP &\nbody count in quantifying the impact of policy-changes & disasters, and help\nmanage electronic idea-streams in ways that strengthen community networks.\nEmpirical methods for acquiring task-layer multiplicity data are in their\ninfancy, although experience-sampling via cell-phone button-clicks might be one\nplace to start.\n", " Boolean networks with canalizing functions are used to model gene regulatory\nnetworks. In order to learn how such networks may behave under evolutionary\nforces, we simulate the evolution of a single Boolean network by means of an\nadaptive walk, which allows us to explore the fitness landscape. Mutations\nchange the connections and the functions of the nodes. Our fitness criterion is\nthe robustness of the dynamical attractors against small perturbations. We find\nthat with this fitness criterion the global maximum is always reached and that\nthere is a huge neutral space of 100% fitness. Furthermore, in spite of having\nsuch a high degree of robustness, the evolved networks still share many\nfeatures with \"chaotic\" networks.\n", " Summary: The ModuLand plug-in provides Cytoscape users an algorithm for\ndetermining extensively overlapping network modules. Moreover, it identifies\nseveral hierarchical layers of modules, where meta-nodes of the higher\nhierarchical layer represent modules of the lower layer. The tool assigns\nmodule cores, which predict the function of the whole module, and determines\nkey nodes bridging two or multiple modules. The plug-in has a detailed\nJAVA-based graphical interface with various colouring options. The ModuLand\ntool can run on Windows, Linux, or Mac OS. We demonstrate its use on protein\nstructure and metabolic networks. Availability: The plug-in and its user guide\ncan be downloaded freely from: http://www.linkgroup.hu/modules.php. Contact:\[email protected] Supplementary information: Supplementary\ninformation is available at Bioinformatics online.\n", " Recent interest in integrated sensing and communications has led to the\ndesign of novel signal processing techniques to recover information from an\noverlaid radar-communications signal. Here, we focus on a spectral coexistence\nscenario, wherein the channels and transmit signals of both radar and\ncommunications systems are unknown to the common receiver. In this dual-blind\ndeconvolution (DBD) problem, the receiver admits a multi-carrier wireless\ncommunications signal that is overlaid with the radar signal reflected off\nmultiple targets. The communications and radar channels are represented by\ncontinuous-valued range-times or delays corresponding to multiple transmission\npaths and targets, respectively. Prior works addressed recovery of unknown\nchannels and signals in this ill-posed DBD problem through atomic norm\nminimization but contingent on individual minimum separation conditions for\nradar and communications channels. In this paper, we provide an optimal joint\nseparation condition using extremal functions from the Beurling-Selberg\ninterpolation theory. Thereafter, we formulate DBD as a low-rank modified\nHankel matrix retrieval and solve it via nuclear norm minimization. We estimate\nthe unknown target and communications parameters from the recovered low-rank\nmatrix using multiple signal classification (MUSIC) method. We show that the\njoint separation condition also guarantees that the underlying Vandermonde\nmatrix for MUSIC is well-conditioned. Numerical experiments validate our\ntheoretical findings.\n", " Semantic communication allows the receiver to know the intention instead of\nthe bit information itself, which is an emerging technique to support real-time\nhuman-machine and machine-to-machine interactions for future wireless\ncommunications. In semantic communications, both transmitter and receiver share\nsome common knowledge, which can be used to extract small-size information at\nthe transmitter and recover the original information at the receiver. Due to\ndifferent design purposes, security issues in semantic communications have two\nunique features compared to standard bit-wise communications. First, an\nattacker in semantic communications considers not only the amount of stolen\ndata but also the meanings of stolen data. Second, an attacker in semantic\ncommunication systems can attack not only semantic information transmission as\ndone in standard communication systems but also attacks machine learning (ML)\nmodels used for semantic information extraction since most of semantic\ninformation is generated using ML based methods. Due to these unique features,\nin this paper, we present an overview on the fundamentals and key challenges in\nthe design of secure semantic communication. We first provide various methods\nto define and extract semantic information. Then, we focus on secure semantic\ncommunication techniques in two areas: information security and semantic ML\nmodel security. For each area, we identify the main problems and challenges.\nThen, we will provide a comprehensive treatment of these problems. In a\nnutshell,this article provides a holistic set of guidelines on how to design\nsecure semantic communication systems over real-world wireless communication\nnetworks.\n", " Dynamic connectivity is a well-studied problem, but so far the most\ncompelling progress has been confined to the edge-update model: maintain an\nunderstanding of connectivity in an undirected graph, subject to edge\ninsertions and deletions. In this paper, we study two more challenging, yet\nequally fundamental problems.\n Subgraph connectivity asks to maintain an understanding of connectivity under\nvertex updates: updates can turn vertices on and off, and queries refer to the\nsubgraph induced by \"on\" vertices. (For instance, this is closer to\napplications in networks of routers, where node faults may occur.)\n We describe a data structure supporting vertex updates in O (m^{2/3})\namortized time, where m denotes the number of edges in the graph. This greatly\nimproves over the previous result [Chan, STOC'02], which required fast matrix\nmultiplication and had an update time of O(m^0.94). The new data structure is\nalso simpler.\n Geometric connectivity asks to maintain a dynamic set of n geometric objects,\nand query connectivity in their intersection graph. (For instance, the\nintersection graph of balls describes connectivity in a network of sensors with\nbounded transmission radius.)\n Previously, nontrivial fully dynamic results were known only for special\ncases like axis-parallel line segments and rectangles. We provide similarly\nimproved update times, O (n^{2/3}), for these special cases. Moreover, we show\nhow to obtain sublinear update bounds for virtually all families of geometric\nobjects which allow sublinear-time range queries, such as arbitrary 2D line\nsegments, d-dimensional simplices, and d-dimensional balls.\n", " The present study focuses on the interaction of gravity waves in the\natmosphere with the tropopause. As the vertical extent of the latter is small\ncompared to the density scale height, wave propagation is described by the\nTaylor-Goldstein equation as derived from the linearised Boussinesq\napproximation. Of particular interest in the construction of gravity wave\nparameterisations for the upper atmosphere are the transmission and reflection\nproperties of the tropopause as these determine the upward fluxes of energy and\nmomentum carried by internal waves.\n A method is presented that decomposes internal waves explicitly into upward\nand downward propagating contributions, thus giving direct access to\ntransmission and reflection coefficients of finite regions of non-uniform\nstratification in a stationary atmosphere. The scheme utilizes a piecewise\nconstant approximation for the background stratification and matches up- and\ndownward propagating plane wave solutions in each layer through physically\nmeaningful coupling conditions. As a result, transmission and reflection\ncoefficients follow immediately.\n", " Networks describing the interaction of the elements that constitute a complex\nsystem grow and develop via a number of different mechanisms, such as the\naddition and deletion of nodes, the addition and deletion of edges, as well as\nthe duplication or fusion of nodes. While each of these mechanisms can have a\ndifferent cause depending on whether the network is biological, technological,\nor social, their impact on the network's structure, as well as its local and\nglobal properties, is similar. This allows us to study how each of these\nmechanisms affects networks either alone or together with the other processes,\nand how they shape the characteristics that have been observed. We study how a\nnetwork's growth parameters impact the distribution of edges in the network,\nhow they affect a network's modularity, and point out that some parameters will\ngive rise to networks that have the opposite tendency, namely to display\nanti-modularity. Within the model we are describing, we can search the space of\npossible networks for parameter sets that generate networks that are very\nsimilar to well-known and well-studied examples, such as the brain of a worm,\nand the network of interactions of the proteins in baker's yeast.\n", " The properties of the known superconductors can be explained with the\ncorrelations dominated superconducting mechanism (CSM). The correlations have\nthe spin correlation, the charge correlation and the spin-charge correlation,\nand their strengths can be described by the related correlation lengths in\ntheir correlation functions. Our evaluation from many superconductors is that\nsuperconductivities occur if both the spin correlation and the charge\ncorrelation are stronger, and the calculation of a Hubbard model showed that\nthe spin-charge correlation may govern superconductivities1. Afterwards, this\nmechanism has led a model which includes various superconductivities and\nmagnetisms, and the relation between superconductivities and magnetisms can be\nunderstood on this model2 (these results have been shown by calculations). This\nmechanism is very practical, for example, to turn a material into a\nsuperconductor or increase the Tc of a superconductor, what we will do is to\nincrease the spin-charge correlation. In this letter, we first describe the\nrelations between the spin-charge correlation, the spin correlation and the\ncharge correlation, take these relations as the basis of constructing a new\nphase diagram, and then classify the known superconductors into various\nsections in this phase diagram. This letter also gives a new explanation about\nthe pressure effect on Tc, the isotope effect on Tc and the pairing symmetry\nwith the CSM.\n", " We examine the equilibrium properties of hot, dilute, non-relativistic\nplasmas. The partition function and density correlation functions of a\nclassical plasma with several species are expressed in terms of a functional\nintegral over electrostatic potential distributions. The leading order,\nfield-theoretic tree approximation automatically includes the effects of Debye\nscreening. Subleading, one-loop corrections are easily evaluated. The two-loop\ncorrections, however, have ultraviolet divergences. These correspond to the\nshort-distance, logarithmic divergence which is encountered in the spatial\nintegral of the Boltzmann exponential when it is expanded to third order in the\nCoulomb potential. Such divergences do not appear in the underlying quantum\ntheory --- they are rendered finite by quantum fluctuations. We show how such\ndivergences may be removed and the correct finite theory obtained by\nintroducing additional local interactions in the manner of modern effective\nquantum field theories. We obtain explicit results for density-density\ncorrelation functions through two-loop order and thermodynamic quantities\nthrough three-loop order. The induced couplings are shown to obey\nrenormalization group equations, and these equations are used to characterize\nall leading logarithmic contributions in the theory. A linear combination of\npressure and energy and number densities is shown to be described by a\nfield-theoretic anomaly. The effective theory allows us to evaluate very easily\nthe algebraic long-distance decay of density correlation functions.\n", " Given a set $P$ of $n$ points in the plane, where $n$ is even, we consider\nthe following question: How many plane perfect matchings can be packed into\n$P$? We prove that at least $\\lceil\\log_2{n}\\rceil-2$ plane perfect matchings\ncan be packed into any point set $P$. For some special configurations of point\nsets, we give the exact answer. We also consider some extensions of this\nproblem.\n", " Astrocytes affect neural transmission by a tight control via glutamate\ntransporters on glutamate concentrations in direct vicinity to the synaptic\ncleft and by extracellular glutamate. Their relevance for information\nrepresentation has been supported by in-vivo studies in ferret and mouse\nprimary visual cortex. In ferret blocking glutamate transport pharmacologically\nbroadened tuning curves and enhanced the response at preferred orientation. In\nknock-out mice with reduced expression of glutamate transporters sharpened\ntuning was observed. It is however unclear how focal and ambient changes in\nglutamate concentration affect stimulus representation. Here we develop a\ncomputational framework, which allows the investigation of synaptic and\nextrasynaptic effects of glutamate uptake on orientation tuning in recurrently\nconnected network models with pinwheel-domain (ferret) or salt-and-pepper\n(mouse) organization. This model proposed that glutamate uptake shapes\ninformation representation when it affects the contribution of excitatory and\ninhibitory neurons to the network activity. Namely, strengthening the\ncontribution of excitatory neurons generally broadens tuning and elevates the\nresponse. In contrast, strengthening the contribution of inhibitory neurons can\nhave a sharpening effect on tuning. In addition local representational topology\nalso plays a role: In the pinwheel-domain model effects were strongest within\ndomains - regions where neighboring neurons share preferred orientations.\nAround pinwheels but also within salt-and-pepper networks the effects were less\nstrong. Our model proposes that the pharmacological intervention in ferret\nincreases the contribution of excitatory cells, while the reduced expression in\nmouse increases the contribution of inhibitory cells to network activity.\n", " Motivated by the need to correct the potentially large kinematic errors in\napproximations used in the standard formulation of perturbative QCD, we\nreformulate deeply inelastic lepton-proton scattering in terms of gauge\ninvariant, universal parton correlation functions which depend on all\ncomponents of parton four-momentum. Currently, different hard QCD processes are\ndescribed by very different perturbative formalisms, each relying on its own\nset of kinematical approximations. In this paper we show how to set up\nformalism that avoids approximations on final-state momenta, and thus has a\nvery general domain of applicability. The use of exact kinematics introduces a\nnumber of significant conceptual shifts already at leading order, and tightly\nconstrains the formalism. We show how to define parton correlation functions\nthat generalize the concepts of parton density, fragmentation function, and\nsoft factor. After setting up a general subtraction formalism, we obtain a\nfactorization theorem. To avoid complications with Ward identities the full\nderivation is restricted to abelian gauge theories; even so the resulting\nstructure is highly suggestive of a similar treatment for non-abelian gauge\ntheories.\n" ]

[ " Moment control of a simple quadratic reaction network describing a\ndimerization process is addressed. It is shown that the moment closure problem\ncan be circumvented without invoking any moment closure technique. Local\nstabilization and convergence of the average dimer population to any desired\nreference value is ensured using a pure integral control law. Explicit bounds\non the controller gain are provided and shown to be valid for any reference\nvalue. As a byproduct, an explicit upper-bound of the variance of the monomer\nspecies, acting on the system as unknown input due to the moment openness, is\nobtained. The obtained results are illustrated by an example relying on the\nsimulation of a cell population using stochastic simulation algorithms.\n", " This paper introduces a novel operator, termed the Y operator, to elevate\ncontrol performance in Actor-Critic(AC) based reinforcement learning for\nsystems governed by stochastic differential equations(SDEs). The Y operator\ningeniously integrates the stochasticity of a class of child-mother system into\nthe Critic network's loss function, yielding substantial advancements in the\ncontrol performance of RL algorithms.Additionally, the Y operator elegantly\nreformulates the challenge of solving partial differential equations for the\nstate-value function into a parallel problem for the drift and diffusion\nfunctions within the system's SDEs.A rigorous mathematical proof confirms the\noperator's validity.This transformation enables the Y Operator-based\nReinforcement Learning(YORL) framework to efficiently tackle optimal control\nproblems in both model-based and data-driven systems.The superiority of YORL is\ndemonstrated through linear and nonlinear numerical examples showing its\nenhanced performance over existing methods post convergence.\n", " We report on a dramatic improvement of the performance of the classical\ntime-delayed autosynchronization method (TDAS) to control unstable steady\nstates, by applying a time-varying delay in the TDAS control scheme in a form\nof a deterministic or stochastic delay-modulation in a fixed interval around a\nnominal value $T_0$. The successfulness of this variable delay feedback control\n(VDFC) is illustrated by a numerical control simulation of the Lorenz and\nR\\\"{o}ssler systems using three different types of time-delay modulations: a\nsawtooth wave, a sine wave, and a uniform random distribution. We perform a\ncomparative analysis between the VDFC method and the standard TDAS method for a\nsawtooth-wave modulation by analytically determining the domains of control for\nthe generic case of an unstable fixed point of focus type.\n", " Learning to control the structure of sentences is a challenging problem in\ntext generation. Existing work either relies on simple deterministic approaches\nor RL-based hard structures. We explore the use of structured variational\nautoencoders to infer latent templates for sentence generation using a soft,\ncontinuous relaxation in order to utilize reparameterization for training.\nSpecifically, we propose a Gumbel-CRF, a continuous relaxation of the CRF\nsampling algorithm using a relaxed Forward-Filtering Backward-Sampling (FFBS)\napproach. As a reparameterized gradient estimator, the Gumbel-CRF gives more\nstable gradients than score-function based estimators. As a structured\ninference network, we show that it learns interpretable templates during\ntraining, which allows us to control the decoder during testing. We demonstrate\nthe effectiveness of our methods with experiments on data-to-text generation\nand unsupervised paraphrase generation.\n", " The study of control mechanisms of biological systems allows for interesting\napplications in bioengineering and medicine, for instance in cell reprogramming\nor drug target identification. A control strategy often consists of a set of\ninterventions that, by fixing the values of some components, ensure that the\nlong term dynamics of the controlled system is in a desired state. A common\napproach to control in the Boolean framework consists in checking how the fixed\nvalues propagate through the network, to establish whether the effect of\npercolating the interventions is sufficient to induce the target state.\nAlthough methods based uniquely on value percolation allow for efficient\ncomputation, they can miss many control strategies. Exhaustive methods for\ncontrol strategy identification, on the other hand, often entail high\ncomputational costs. In order to increase the number of control strategies\nidentified while still benefiting from an efficient implementation, we\nintroduce a method based on value percolation that uses trap spaces, subspaces\nof the state space that are closed with respect to the dynamics, and that can\nusually be easily computed in biological networks. The approach allows for node\ninterventions, which fix the value of certain components, and edge\ninterventions, which fix the effect that one component has on another. The\nmethod is implemented using Answer Set Programming, extending an existing\nefficient implementation of value percolation to allow for the use of trap\nspaces and edge control. The applicability of the approach is studied for\ndifferent control targets in a biological case study, identifying in all cases\nnew control strategies that would escape usual percolation-based methods.\n", " Chemical reaction networks (CRNs) provide a convenient language for modelling\na broad variety of biological systems. These models are commonly studied with\nrespect to the time series they generate in deterministic or stochastic\nsimulations. Their dynamic behaviours are then analysed, often by using\ndeterministic methods based on differential equations with a focus on the\nsteady states. Here, we propose a method for comparing CRNs with respect to\ntheir behaviour in stochastic simulations. Our method is based on using the\nflux graphs that are delivered by stochastic simulations as abstract\nrepresentations of their dynamic behaviour. This allows us to compare the\nbehaviour of any two CRNs for any time interval, and define a notion of\nequivalence on them that overlaps with graph isomorphism at the lowest level of\nrepresentation. The similarity between the compared CRNs can be quantified in\nterms of their distance. The results can then be used to refine the models or\nto replace a larger model with a smaller one that produces the same behaviour\nor vice versa.\n", " We consider a simple control problem in which the underlying dynamics depend\non a parameter $a$ that is unknown and must be learned. We study three variants\nof the control problem: Bayesian control, in which we have a prior belief about\n$a$; bounded agnostic control, in which we have no prior belief about $a$ but\nwe assume that $a$ belongs to a bounded set; and fully agnostic control, in\nwhich $a$ is allowed to be an arbitrary real number about which we have no\nprior belief. In the Bayesian variant, a control strategy is optimal if it\nminimizes a certain expected cost. In the agnostic variants, a control strategy\nis optimal if it minimizes a quantity called the worst-case regret. For the\nBayesian and bounded agnostic variants above, we produce optimal control\nstrategies. For the fully agnostic variant, we produce almost optimal control\nstrategies, i.e., for any $\\varepsilon>0$ we produce a strategy that minimizes\nthe worst-case regret to within a multiplicative factor of $(1+\\varepsilon)$.\n", " Stochastic kinetic models of genetic expression are able to describe protein\nfluctuations. A comparative study of the canonical and a feedback model is\ngiven here by using stochastic simulation methods. The feedback model is\nskeleton model implementation of the circular gene hypothesis, which suggests\nthe interaction between the synthesis and degradation of mRNA. Qualitative and\nquantitative changes in the shape and in the numerical characteristics of the\nstationary distributions suggest that more combined experimental and\ntheoretical studies should be done to uncover the details of the kinetic\nmechanism of gene expression.\n", " The probability distribution describing the state of a Stochastic Reaction\nNetwork evolves according to the Chemical Master Equation (CME). It is common\nto estimated its solution using Monte Carlo methods such as the Stochastic\nSimulation Algorithm (SSA). In many cases these simulations can take an\nimpractical amount of computational time. Therefore many methods have been\ndeveloped that approximate the Stochastic Process underlying the Chemical\nMaster Equation. Prominent strategies are Hybrid Models that regard the firing\nof some reaction channels as being continuous and applying the quasi-stationary\nassumption to approximate the dynamics of fast subnetworks. However as the\ndynamics of a Stochastic Reaction Network changes with time these\napproximations might have to be adapted during the simulation. We develop a\nmethod that approximates the solution of a CME by automatically partitioning\nthe reaction dynamics into discrete/continuous components and applying the\nquasi-stationary assumption on identifiable fast subnetworks. Our method does\nnot require user intervention and it adapts to exploit the changing timescale\nseparation between reactions and/or changing magnitudes of copy numbers of\nconstituent species. We demonstrate the efficiency of the proposed method by\nconsidering examples from Systems Biology and showing that very good\napproximations to the exact probability distributions can be achieved in\nsignificantly less computational time.\n", " Understanding control mechanisms in biological systems plays a crucial role\nin important applications, for instance in cell reprogramming. Boolean modeling\nallows the identification of possible efficient strategies, helping to reduce\nthe usually high and time-consuming experimental efforts. Available approaches\nto control strategy identification usually focus either on attractor or\nphenotype control, and are unable to deal with more complex control problems,\nfor instance phenotype avoidance. They also fail to capture, in many\nsituations, all possible minimal strategies, finding instead only sub-optimal\nsolutions. In order to fill these gaps, we present a novel approach to control\nstrategy identification in Boolean networks based on model checking. The method\nis guaranteed to identify all minimal control strategies, and provides maximal\nflexibility in the definition of the control target. We investigate the\napplicability of the approach by considering a range of control problems for\ndifferent biological systems, comparing the results, where possible, to those\nobtained by alternative control methods.\n", " This paper analyzes, in the context of a prokaryotic cell, the stochastic\nvariability of the number of proteins when there is a control of gene\nexpression by an autoregulation scheme. The goal of this work is to estimate\nthe efficiency of the regulation to limit the fluctuations of the number of\ncopies of a given protein. The autoregulation considered in this paper relies\nmainly on a negative feedback: the proteins are repressors of their own gene\nexpression. The efficiency of a production process without feedback control is\ncompared to a production process with an autoregulation of the gene expression\nassuming that both of them produce the same average number of proteins. The\nmain characteristic used for the comparison is the standard deviation of the\nnumber of proteins at equilibrium. With a Markovian representation and a simple\nmodel of repression, we prove that, under a scaling regime, the repression\nmechanism follows a Hill repression scheme with an hyperbolic control. An\nexplicit asymptotic expression of the variance of the number of proteins under\nthis regulation mechanism is obtained. Simulations are used to study other\naspects of autoregulation such as the rate of convergence to equilibrium of the\nproduction process and the case where the control of the production process of\nproteins is achieved via the inhibition of mRNAs.\n", " We introduce a framework for calibrating machine learning models so that\ntheir predictions satisfy explicit, finite-sample statistical guarantees. Our\ncalibration algorithms work with any underlying model and (unknown)\ndata-generating distribution and do not require model refitting. The framework\naddresses, among other examples, false discovery rate control in multi-label\nclassification, intersection-over-union control in instance segmentation, and\nthe simultaneous control of the type-1 error of outlier detection and\nconfidence set coverage in classification or regression. Our main insight is to\nreframe the risk-control problem as multiple hypothesis testing, enabling\ntechniques and mathematical arguments different from those in the previous\nliterature. We use the framework to provide new calibration methods for several\ncore machine learning tasks, with detailed worked examples in computer vision\nand tabular medical data.\n", " Signaling in enzymatic networks is typically triggered by environmental\nfluctuations, resulting in a series of stochastic chemical reactions, leading\nto corruption of the signal by noise. For example, information flow is\ninitiated by binding of extracellular ligands to receptors, which is\ntransmitted through a {cascade involving} kinase-phosphatase stochastic\nchemical reactions. For a class of such networks, we develop a general\nfield-theoretic approach in order to calculate the error in signal transmission\nas a function of an appropriate control variable. Application of the theory to\na simple push-pull network, a module in the kinase-phosphatase cascade,\nrecovers the exact results for error in signal transmission previously obtained\nusing umbral calculus (Phys. Rev. X., {\\bf 4}, 041017 (2014)). We illustrate\nthe generality of the theory by studying the minimal errors in noise reduction\nin a reaction cascade with two connected push-pull modules. Such a cascade\nbehaves as an effective {three-species} network with a pseudo intermediate. In\nthis case, optimal information transfer, resulting in the smallest square of\nthe error between the input and output, occurs with a time delay, which is\ngiven by the inverse of the decay rate of the pseudo intermediate. There are\nsubstantial deviations between simulations and predictions of the linear theory\nin error in signal propagation in an enzymatic push-pull network for a certain\nrange of parameters. Inclusion of second order perturbative corrections shows\nthat differences between simulations and theoretical predictions are minimized.\nOur study establishes that {a} field theoretic formulation {of} stochastic\nbiological {signaling offers} a systematic way to understand error propagation\nin networks of arbitrary complexity.\n", " Exploiting recent developments in information theory, we propose, illustrate,\nand validate a principled information-theoretic algorithm for module discovery\nand resulting measure of network modularity. This measure is an order parameter\n(a dimensionless number between 0 and 1). Comparison is made to other\napproaches to module-discovery and to quantifying network modularity using\nMonte Carlo generated Erdos-like modular networks. Finally, the Network\nInformation Bottleneck (NIB) algorithm is applied to a number of real world\nnetworks, including the \"social\" network of coauthors at the APS March Meeting\n2004.\n", " We present a new, very short proof of a conjecture by I. Ra\\c{s}a, which is\nan inequality involving basic Bernstein polynomials and convex functions. It\nwas affirmed positively very recently by J. Mrowiec, T. Rajba and S.\nW\\k{a}sowicz (2017) by the use of stochastic convex orders, as well as by Abel\n(2017) who simplified their proof. We give a useful sufficient condition for\nthe verification of some stochastic convex order relations, which in the case\nof binomial distributions are equivalent to the I. Ra\\c{s}a inequality. We give\nalso the corresponding inequalities for other distributions. Our methods allow\nus to give some extended versions of stochastic convex orderings as well as the\nI. Ra\\c{s}a type inequalities. In particular, we prove the Muirhead type\ninequality for convex orders for convolution polynomials of probability\ndistributions.\n", " We give a short survey on computational techniques which can be used to solve\nthe representation conversion problem for polyhedra up to symmetries. We in\nparticular discuss decomposition methods, which reduce the problem to a number\nof lower dimensional subproblems. These methods have been successfully used by\ndifferent authors in special contexts. Moreover, we sketch an incremental\nmethod, which is a generalization of Fourier-Motzkin elimination, and we give\nsome ideas how symmetry can be exploited using pivots.\n", " Optimizing behaviors for dexterous manipulation has been a longstanding\nchallenge in robotics, with a variety of methods from model-based control to\nmodel-free reinforcement learning having been previously explored in\nliterature. Perhaps one of the most powerful techniques to learn complex\nmanipulation strategies is imitation learning. However, collecting and learning\nfrom demonstrations in dexterous manipulation is quite challenging. The\ncomplex, high-dimensional action-space involved with multi-finger control often\nleads to poor sample efficiency of learning-based methods. In this work, we\npropose 'Dexterous Imitation Made Easy' (DIME) a new imitation learning\nframework for dexterous manipulation. DIME only requires a single RGB camera to\nobserve a human operator and teleoperate our robotic hand. Once demonstrations\nare collected, DIME employs standard imitation learning methods to train\ndexterous manipulation policies. On both simulation and real robot benchmarks\nwe demonstrate that DIME can be used to solve complex, in-hand manipulation\ntasks such as 'flipping', 'spinning', and 'rotating' objects with the Allegro\nhand. Our framework along with pre-collected demonstrations is publicly\navailable at https://nyu-robot-learning.github.io/dime.\n", " Neuroscientific data analysis has classically involved methods for\nstatistical signal and image processing, drawing on linear algebra and\nstochastic process theory. However, digitized neuroanatomical data sets\ncontaining labelled neurons, either individually or in groups labelled by\ntracer injections, do not fully fit into this classical framework. The\ntree-like shapes of neurons cannot mathematically be adequately described as\npoints in a vector space. There is therefore a need for new approaches. Methods\nfrom computational topology and geometry are naturally suited to the analysis\nof neuronal shapes. Here we introduce methods from Discrete Morse Theory to\nextract tree-skeletons of individual neurons from volumetric brain image data,\nor to summarize collections of neurons labelled by localized anterograde tracer\ninjections. Since individual neurons are topologically trees, it is sensible to\nsummarize the collection of neurons labelled by a localized anterograde tracer\ninjection using a consensus tree-shape. The algorithmic procedure includes an\ninitial pre-processing step to extract a density field from the raw volumetric\nimage data, followed by initial skeleton extraction from the density field\nusing a discrete version of a 1-(un)stable manifold of the density field.\nHeuristically, if the density field is regarded as a mountainous landscape,\nthen the 1-(un)stable manifold follows the \"mountain ridges\" connecting the\nmaxima of the density field. We then simplify this skeleton-graph into a tree\nusing a shortest-path approach and methods derived from persistent homology.\nThe advantage of this approach is that it uses global information about the\ndensity field and is therefore robust to local fluctuations and non-uniformly\ndistributed input signals. To be able to handle large data sets, we use a\ndivide-and-conquer approach. The resulting software DiMorSC is available on\nGithub.\n", " In this paper, we consider the problem of existence and uniqueness of\nsolutions to the Einstein field equations for a spatially flat FLRW universe in\nthe context of stochastic eternal inflation where the stochastic mechanism is\nmodelled by adding a stochastic forcing term representing Gaussian white noise\nto the Klein-Gordon equation. We show that under these considerations, the\nKlein-Gordon equation actually becomes a stochastic differential equation.\nTherefore, the existence and uniqueness of solutions to Einstein's equations\ndepend on whether the coefficients of this stochastic differential equation\nobey Lipschitz continuity conditions. We show that for any choice of $V(\\phi)$,\nthe Einstein field equations are not globally well-posed, hence, any solution\nfound to these equations is not guaranteed to be unique. Instead, the\ncoefficients are at best locally Lipschitz continuous in the physical state\nspace of the dynamical variables, which only exist up to a finite explosion\ntime. We further perform Feller's explosion test for an arbitrary power-law\ninflaton potential and prove that all solutions to the EFE explode in a finite\ntime with probability one. This implies that the mechanism of stochastic\ninflation thus considered cannot be described to be eternal, since the very\nconcept of eternal inflation implies that the process continues indefinitely.\nWe therefore argue that stochastic inflation based on a stochastic forcing term\nwould not produce an infinite number of universes in some multiverse ensemble.\nIn general, since the Einstein field equations in both situations are not\nwell-posed, we further conclude that the existence of a multiverse via the\nstochastic eternal inflation mechanism considered in this paper is still very\nmuch an open question that will require much deeper investigation.\n", " We address the problem of tracking and detecting interactions between the\ndifferent groups of runners that form during a race. In athletic races control\npoints are set to monitor the progress of athletes over the course.\nIntuitively, a {\\it group} is a sufficiently large set of athletes that cross a\ncontrol point together. After adapting an existing definition of group to our\nsetting we go on to study two types of group evolution patterns. The primary\nfocus of this work are {\\it evolution patterns}, i.e. the transformation and\ninteraction of groups of athletes between two consecutive control points. We\nprovide an accurate geometric model of the following evolution patterns:\nsurvives, appears, disappears, expands, shrinks, merges, splits, coheres and\ndisbands, and present algorithms to efficiently compute these patterns. Next,\nbased on the algorithms introduced for identifying evolution patterns,\nalgorithms to detect {\\it long-term patterns} are introduced. These patterns\ntrack global properties over several control points: surviving, traceable\nforward, traceable backward and related forward and backward. Experimental\nevaluation of the algorithms provided is presented using real and synthetic\ndata. Using the data currently available, our experiments show how our\nalgorithms can provide valuable insight into how running races develop.\nMoreover, we also show how, even if dense (synthetic) data is considered, our\nalgorithms are also able to process it in real time.\n", " This paper addresses the problem of statistical inference for latent\ncontinuous-time stochastic processes, which is often intractable, particularly\nfor discrete state space processes described by Markov jump processes. To\novercome this issue, we propose a new tractable inference scheme based on an\nentropic matching framework that can be embedded into the well-known\nexpectation propagation algorithm. We demonstrate the effectiveness of our\nmethod by providing closed-form results for a simple family of approximate\ndistributions and apply it to the general class of chemical reaction networks,\nwhich are a crucial tool for modeling in systems biology. Moreover, we derive\nclosed form expressions for point estimation of the underlying parameters using\nan approximate expectation maximization procedure. We evaluate the performance\nof our method on various chemical reaction network instantiations, including a\nstochastic Lotka-Voltera example, and discuss its limitations and potential for\nfuture improvements. Our proposed approach provides a promising direction for\naddressing complex continuous-time Bayesian inference problems.\n", " It is an outstanding problem whether a pipe-flow system on a star-shaped\nnetwork is stabilisable by a feedback control on the common vertex. In the\npresent paper we deal with this problem. In particular, we study the equation\ngoverning the small vibrations of a stretched elastic pipe conveying fluid in a\nstar-shaped network and examine the question of vertex feedback stabilisability\nof such a system via control moments. Finding an answer to the question is not\nstraightforward, for the system operator associated with the corresponding\nclosed-loop system is unbounded and nonselfadjoint. An approach to the study of\nthe stabilisation problem for the closed-loop system is presented based on the\nspectral approach previously introduced by the authors for star graphs of\nstretched elastic beams. When the tension in the pipes is greater than the\nsquare of the fluid-flow velocity, we establish a positive result that in fact\ngives the strong property of uniform exponential stability of the closed-loop\nsystem.\n", " It is not until we become senior citizens do we recognise how much we took\nmaintaining a simple standing posture for granted. It is truly fascinating to\nobserve the magnitude of control the human brain exercises, in real time, to\nactivate and deactivate the lower body muscles and solve a multi-link 3D\ninverted pendulum problem in order to maintain a stable standing posture. This\nrealisation is even more apparent when training an artificial intelligence (AI)\nagent to maintain a standing posture of a digital musculoskeletal avatar due to\nthe error propagation problem. In this work we address the error propagation\nproblem by introducing an iterative training procedure for deep reinforcement\nlearning which allows the agent to learn a finite set of actions and how to\ncoordinate between them in order to achieve a stable standing posture. The\nproposed training approach allowed the agent to increase standing duration from\n4 seconds using the traditional training method to 348 seconds using the\nproposed method. The proposed training method allowed the agent to generalise\nand accommodate perception and actuation noise for almost 108 seconds.\n", " Transfinite patches provide a simple and elegant solution to the problem of\nrepresenting non-four-sided continuous surfaces, which are useful in a variety\nof applications, such as curve network based design. Real-time responsiveness\nis essential in this context, and thus reducing the computation cost is an\nimportant concern. The Midpoint Coons (MC) patch presented in this paper is a\nfusion of two previous transfinite schemes, combining the speed of one with the\nsuperior control mechanism of the other. This is achieved using a new\nconstrained parameterization based on generalized barycentric coordinates and\ntransfinite blending functions.\n", " In real-world scenarios, it is appealing to learn a model carrying out\nstochastic operations internally, known as stochastic computation graphs\n(SCGs), rather than learning a deterministic mapping. However, standard\nbackpropagation is not applicable to SCGs. We attempt to address this issue\nfrom the angle of cost propagation, with local surrogate costs, called\nQ-functions, constructed and learned for each stochastic node in an SCG. Then,\nthe SCG can be trained based on these surrogate costs using standard\nbackpropagation. We propose the entire framework as a solution to generalize\nbackpropagation for SCGs, which resembles an actor-critic architecture but\nbased on a graph. For broad applicability, we study a variety of SCG structures\nfrom one cost to multiple costs. We utilize recent advances in reinforcement\nlearning (RL) and variational Bayes (VB), such as off-policy critic learning\nand unbiased-and-low-variance gradient estimation, and review them in the\ncontext of SCGs. The generalized backpropagation extends transported learning\nsignals beyond gradients between stochastic nodes while preserving the benefit\nof backpropagating gradients through deterministic nodes. Experimental\nsuggestions and concerns are listed to help design and test any specific model\nusing this framework.\n", " We introduce a new class of fully nonlinear integro-differential operators\nwith possible nonsymmetric kernels, which includes the ones that arise from\nstochastic control problems with purely jump L\\`evy processes. If the index of\nthe operator $\\sigma$ is in $ (1,2)$ (subcritical case), then we obtain a\ncomparison principle, a nonlocal version of the Alexandroff-Backelman-Pucci\nestimate, a Harnack inequality, a H\\\"older regularity, and an interior $\\rm\nC^{1,\\alpha}$-regularity for fully nonlinear integro-differential equations\nassociated with such a class. Moreover, our estimates remain uniform as the\nindex $\\sigma$ of the operator is getting close to two, so that they can be\nregarded as a natural extension of regularity results for elliptic partial\ndifferential equations.\n", " The rational covariance extension problem (RCEP) is an important problem in\nsystems and control occurring in such diverse fields as control, estimation,\nsystem identification, and signal and image processing, leading to many\nfundamental theoretical questions. In fact, this inverse problem is a key\ncomponent in many identification and signal processing techniques and plays a\nfundamental role in prediction, analysis, and modeling of systems and signals.\nIt is well-known that the RCEP can be reformulated as a (truncated)\ntrigonometric moment problem subject to a rationality condition. In this paper\nwe consider the more general multidimensional trigonometric moment problem with\na similar rationality constraint. This generalization creates many interesting\nnew mathematical questions and also provides new insights into the original\none-dimensional problem. A key concept in this approach is the complete smooth\nparametrization of all solutions, allowing solutions to be tuned to satisfy\nadditional design specifications without violating the complexity constraints.\nAs an illustration of the potential of this approach we apply our results to\nmultidimensional spectral estimation and image compression. This is just a\nfirst step in this direction, and we expect that more elaborate tuning\nstrategies will enhance our procedures in the future.\n", " Reinforcement learning (RL) has shown great promise in optimizing long-term\nuser interest in recommender systems. However, existing RL-based recommendation\nmethods need a large number of interactions for each user to learn a robust\nrecommendation policy. The challenge becomes more critical when recommending to\nnew users who have a limited number of interactions. To that end, in this\npaper, we address the cold-start challenge in the RL-based recommender systems\nby proposing a meta-level model-based reinforcement learning approach for fast\nuser adaptation. In our approach, we learn to infer each user's preference with\na user context variable that enables recommendation systems to better adapt to\nnew users with few interactions. To improve adaptation efficiency, we learn to\nrecover the user policy and reward from only a few interactions via an inverse\nreinforcement learning method to assist a meta-level recommendation agent.\nMoreover, we model the interaction relationship between the user model and\nrecommendation agent from an information-theoretic perspective. Empirical\nresults show the effectiveness of the proposed method when adapting to new\nusers with only a single interaction sequence. We further provide a theoretical\nanalysis of the recommendation performance bound.\n", " Inspired by the seminal result that a graph and an associated rotation system\nuniquely determine the topology of a closed manifold, we propose a\ncombinatorial method for reconstruction of surfaces from points. Our method\nconstructs a spanning tree and a rotation system. Since the tree is trivially a\nplanar graph, its rotation system determines a genus zero surface with a single\nface which we proceed to incrementally refine by inserting edges to split faces\nand thus merging them. In order to raise the genus, special handles are added\nby inserting edges between different faces and thus merging them. We apply our\nmethod to a wide range of input point clouds in order to investigate its\neffectiveness, and we compare our method to several other surface\nreconstruction methods. We find that our method offers better control over\noutlier classification, i.e. which points to include in the reconstructed\nsurface, and also more control over the topology of the reconstructed surface.\n", " Stochastic Boolean networks, or more generally, stochastic discrete networks,\nare an important class of computational models for molecular interaction\nnetworks. The stochasticity stems from the updating schedule. Standard updating\nschedules include the synchronous update, where all the nodes are updated at\nthe same time, and the asynchronous update where a random node is updated at\neach time step. The former produces a deterministic dynamics while the latter a\nstochastic dynamics. A more general stochastic setting considers propensity\nparameters for updating each node. Stochastic Discrete Dynamical Systems (SDDS)\nis a modeling framework that considers two propensity parameters for updating\neach node and uses one when the update has a positive impact on the variable,\nthat is, when the update causes the variable to increase its value, and uses\nthe other when the update has a negative impact, that is, when the update\ncauses it to decrease its value. This framework offers additional features for\nsimulations but also adds a complexity in parameter estimation of the\npropensities. This paper presents a method for estimating the propensity\nparameters for SDDS. The method is based on adding noise to the system using\nthe Google PageRank approach to make the system ergodic and thus guaranteeing\nthe existence of a stationary distribution. Then with the use of a genetic\nalgorithm, the propensity parameters are estimated. Approximation techniques\nthat make the search algorithms efficient are also presented and Matlab/Octave\ncode to test the algorithms are available at\nhttp://www.ms.uky.edu/~dmu228/GeneticAlg/Code.html.\n", " Intrinsic noise, the stochastic cell-to-cell fluctuations in mRNAs and\nproteins, has been observed and proved to play important roles in cellular\nsystems. Due to the recent development in single-cell-level measurement\ntechnology, the studies on intrinsic noise are becoming increasingly popular\namong scholars. The chemical master equation (CME) has been used to model the\nevolutions of complex chemical and biological systems since 1940, and are often\nserved as the standard tool for modeling intrinsic noise in gene regulation\nsystem. A CME-based model can capture the discrete, stochastic, and dynamical\nnature of gene regulation system, and may offer casual and physical explanation\nof the observed data at single-cell level. Nonetheless, the complexity of CME\nalso pose serious challenge for researchers in proposing practical modeling and\ninference frameworks. In this article, we will review the existing works on the\nmodelings and inference of intrinsic noise in gene regulation system within the\nframework of CME model. We will explore the principles in constructing a CME\nmodel for studying gene regulation system and discuss the popular\napproximations of CME. Then we will study the simulation simulation methods as\nwell as the analytical and numerical approaches that can be used to obtain\nsolution to a CME model. Finally we will summary the exiting statistical\nmethods that can be used to infer the unknown parameters or structures in CME\nmodel using single-cell-level gene expression data.\n", " In many stochastic simulations of biochemical reaction networks, it is\ndesirable to ``coarse-grain'' the reaction set, removing fast reactions while\nretaining the correct system dynamics. Various coarse-graining methods have\nbeen proposed, but it remains unclear which methods are reliable and which\nreactions can safely be eliminated. We address these issues for a model gene\nregulatory network that is particularly sensitive to dynamical fluctuations: a\nbistable genetic switch. We remove protein-DNA and/or protein-protein\nassociation-dissociation reactions from the reaction set, using various\ncoarse-graining strategies. We determine the effects on the steady-state\nprobability distribution function and on the rate of fluctuation-driven switch\nflipping transitions. We find that protein-protein interactions may be safely\neliminated from the reaction set, but protein-DNA interactions may not. We also\nfind that it is important to use the chemical master equation rather than\nmacroscopic rate equations to compute effective propensity functions for the\ncoarse-grained reactions.\n", " Understanding relationship between noisy dynamics and biological network\narchitecture is a fundamentally important question, particularly in order to\nelucidate how cells encode and process information. We analytically and\nnumerically investigate general network architectural conditions that are\nnecessary to generate stochastic amplified and coherent oscillations. We\nenumerate all possible topologies of coupled negative feedbacks in the\nunderlying biochemical networks with three components, negative feedback loops,\nand mass action kinetics. Using the linear noise approximation to analytically\nobtain the time-dependent solution of the master equation and derive the\nalgebraic expression of power spectra, we find that (a) all networks with\ncoupled negative feedbacks are capable of generating stochastic amplified and\ncoherent oscillations; (b) networks with a single negative feedback are better\nstochastic amplified and coherent oscillators than those with multiple coupled\nnegative feedbacks; (c) multiple timescale difference among the kinetic rate\nconstants is required for stochastic amplified and coherent oscillations.\n", " The alternate row and column scaling algorithm applied to a positive $n\\times\nn$ matrix $A$ converges to a doubly stochastic matrix $S(A)$, sometimes called\nthe \\emph{Sinkhorn limit} of $A$. For every positive integer $n$, a two\nparameter family of row but not column stochastic $n\\times n$ positive matrices\nis constructed that become doubly stochastic after exactly one column scaling.\n", " We give a complete characterization of limiting interpolation spa\\-ces for\nthe real method of interpolation using extrapolation theory. For this purpose\nthe usual tools (e.g., Boyd indices or the boundedness of Hardy type operators)\nare not appropriate. Instead, our characterization hinges upon the boundedness\nof some simple operators (e.g. $f\\mapsto f(t^{2})/t$, or $f\\mapsto f(t^{1/2}%\n)$) acting on the underlying lattices that are used to control the $K$- and\n$J$-functionals. Reiteration formulae, extending Holmstedt's classical\nreiteration theorem to limiting spaces, are also proved and characterized in\nthis fashion. The resulting theory gives a unified roof to a large body of\nliterature that, using ad-hoc methods, had covered only special cases of the\nresults obtained here. Applications to Matsaev ideals, Grand Lebesgue spaces,\nBourgain-Brezis-Mironescu-Maz'ya-Shaposhnikova limits, as well as a new vector\nvalued extrapolation theorems, are provided.\n", " We consider the problem of inferring the unknown parameters of a stochastic\nbiochemical network model from a single measured time-course of the\nconcentration of some of the involved species. Such measurements are available,\ne.g., from live-cell fluorescence microscopy in image-based systems biology. In\naddition, fluctuation time-courses from, e.g., fluorescence correlation\nspectroscopy provide additional information about the system dynamics that can\nbe used to more robustly infer parameters than when considering only mean\nconcentrations. Estimating model parameters from a single experimental\ntrajectory enables single-cell measurements and quantification of cell--cell\nvariability. We propose a novel combination of an adaptive Monte Carlo sampler,\ncalled Gaussian Adaptation, and efficient exact stochastic simulation\nalgorithms that allows parameter identification from single stochastic\ntrajectories. We benchmark the proposed method on a linear and a non-linear\nreaction network at steady state and during transient phases. In addition, we\ndemonstrate that the present method also provides an ellipsoidal volume\nestimate of the viable part of parameter space and is able to estimate the\nphysical volume of the compartment in which the observed reactions take place.\n", " In model-based reinforcement learning (MBRL), Wan et al. (2019) showed\nconditions under which the environment model could produce the expectation of\nthe next feature vector rather than the full distribution, or a sample thereof,\nwith no loss in planning performance. Such expectation models are of interest\nwhen the environment is stochastic and non-stationary, and the model is\napproximate, such as when it is learned using function approximation. In these\ncases a full distribution model may be impractical and a sample model may be\neither more expensive computationally or of high variance. Wan et al.\nconsidered only planning for prediction to evaluate a fixed policy. In this\npaper, we treat the control case - planning to improve and find a good\napproximate policy. We prove that planning with an expectation model must\nupdate a state-value function, not an action-value function as previously\nsuggested (e.g., Sorg & Singh, 2010). This opens the question of how planning\ninfluences action selections. We consider three strategies for this and present\ngeneral MBRL algorithms for each. We identify the strengths and weaknesses of\nthese algorithms in computational experiments. Our algorithms and experiments\nare the first to treat MBRL with expectation models in a general setting.\n", " This paper identifies certain interesting mathematical problems of stochastic\nquantization type in the modeling of Laser propagation through turbulent media.\nIn some of the typical physical contexts the problem reduces to stochastic\nSchrodinger equation with space-time white noise of Gaussian, Poisson and Levy\ntype. We identify their mathematical resolution via stochastic quantization.\nNonlinear phenomena such as Kerr effect can be modeled by stochastic nonlinear\nSchrodinger equation in the focusing case with space-time white noise. A\ntreatment of stochastic transport equation, the Korteweg-de Vries Equation as\nwell as a number of other nonlinear wave equations with space-time white noise\nis also given. Main technique is the S-transform (we will actually use closely\nrelated Hermite transform) which converts the stochastic partial differential\nequation with space time white noise to a deterministic partial differential\nequation defined on the Hida-Kondratiev white noise distribution space. We then\nutlize the inverse S-transform/Hermite transform known as the characterization\ntheorem combined with the infinite dimensional implicit function theorem for\nanalytic maps to establish local existence and uniqueness theorems for pathwise\nsolutions of these class of problems. The particular focus of this paper on\nsingular white noise distributions is motivated by practical situations where\nthe refractive index fluctuations in propagation medium in space and time are\nintense due to turbulence, ionospheric plasma turbulence, marine-layer\nfluctuations, etc. Since a large class of partial differential equations that\narise in nonlinear wave propagation have polynomial type nonlinearities, white\nnoise distribution theory is an effective tool in studying these problems\nsubject to different types of white noises.\n", " Estimating fundamental matrices is a classic problem in computer vision.\nTraditional methods rely heavily on the correctness of estimated key-point\ncorrespondences, which can be noisy and unreliable. As a result, it is\ndifficult for these methods to handle image pairs with large occlusion or\nsignificantly different camera poses. In this paper, we propose novel neural\nnetwork architectures to estimate fundamental matrices in an end-to-end manner\nwithout relying on point correspondences. New modules and layers are introduced\nin order to preserve mathematical properties of the fundamental matrix as a\nhomogeneous rank-2 matrix with seven degrees of freedom. We analyze performance\nof the proposed models using various metrics on the KITTI dataset, and show\nthat they achieve competitive performance with traditional methods without the\nneed for extracting correspondences.\n", " Traffic signal control (TSC) is a challenging problem within intelligent\ntransportation systems and has been tackled using multi-agent reinforcement\nlearning (MARL). While centralized approaches are often infeasible for\nlarge-scale TSC problems, decentralized approaches provide scalability but\nintroduce new challenges, such as partial observability. Communication plays a\ncritical role in decentralized MARL, as agents must learn to exchange\ninformation using messages to better understand the system and achieve\neffective coordination. Deep MARL has been used to enable inter-agent\ncommunication by learning communication protocols in a differentiable manner.\nHowever, many deep MARL communication frameworks proposed for TSC allow agents\nto communicate with all other agents at all times, which can add to the\nexisting noise in the system and degrade overall performance. In this study, we\npropose a communication-based MARL framework for large-scale TSC. Our framework\nallows each agent to learn a communication policy that dictates \"which\" part of\nthe message is sent \"to whom\". In essence, our framework enables agents to\nselectively choose the recipients of their messages and exchange variable\nlength messages with them. This results in a decentralized and flexible\ncommunication mechanism in which agents can effectively use the communication\nchannel only when necessary. We designed two networks, a synthetic $4 \\times 4$\ngrid network and a real-world network based on the Pasubio neighborhood in\nBologna. Our framework achieved the lowest network congestion compared to\nrelated methods, with agents utilizing $\\sim 47-65 \\%$ of the communication\nchannel. Ablation studies further demonstrated the effectiveness of the\ncommunication policies learned within our framework.\n", " We present a new method, Non-Stationary Forward Flux Sampling, that allows\nefficient simulation of rare events in both stationary and non-stationary\nstochastic systems. The method uses stochastic branching and pruning to achieve\nuniform sampling of trajectories in phase space and time, leading to accurate\nestimates for time-dependent switching propensities and time-dependent phase\nspace probability densities. The method is suitable for equilibrium or\nnon-equilibrium systems, in or out of stationary state, including non-Markovian\nor externally driven systems. We demonstrate the validity of the technique by\napplying it to a one-dimensional barrier crossing problem that can be solved\nexactly, and show its usefulness by applying it to the time-dependent switching\nof a genetic toggle switch.\n", " I wrote this paper because technology can really improve people's lives. With\nit, we can live longer in a healthy body, save time through increased\nefficiency and automation, and make better decisions. To get to the next level,\nwe need to start looking at intelligence from a much broader perspective, and\npromote international interdisciplinary collaborations. Section 1 of this paper\ndelves into sociology and social psychology to explain that the mechanisms\nunderlying intelligence are inherently social. Section 2 proposes a method to\nclassify intelligence, and describes the differences between weak and strong\nintelligence. Section 3 examines the Chinese Room argument from a different\nperspective. It demonstrates that a Turing-complete machine cannot have strong\nintelligence, and considers the modifications necessary for a computer to be\nintelligent and have understanding. Section 4 argues that the existential risk\ncaused by the technological explosion of a single agent should not be of\nserious concern. Section 5 looks at the AI control problem and argues that it\nis impossible to build a super-intelligent machine that will do what it\ncreators want. By using insights from biology, it also proposes a solution to\nthe control problem. Section 6 discusses some of the implications of strong\nintelligence. Section 7 lists the main challenges with deep learning, and\nasserts that radical changes will be required to reach strong intelligence.\nSection 8 examines a neuroscience framework that could help explain how a\ncortical column works. Section 9 lays out the broad strokes of a road map\ntowards strong intelligence. Finally, section 10 analyzes the impacts and the\nchallenges of greater intelligence.\n", " In this work, we develop a framework for shape analysis using inconsistent\nsurface mapping. Traditional landmark-based geometric morphometrics methods\nsuffer from the limited degrees of freedom, while most of the more advanced\nnon-rigid surface mapping methods rely on a strong assumption of the global\nconsistency of two surfaces. From a practical point of view, given two\nanatomical surfaces with prominent feature landmarks, it is more desirable to\nhave a method that automatically detects the most relevant parts of the two\nsurfaces and finds the optimal landmark-matching alignment between those parts,\nwithout assuming any global 1-1 correspondence between the two surfaces. Our\nmethod is capable of solving this problem using inconsistent surface\nregistration based on quasi-conformal theory. It further enables us to quantify\nthe dissimilarity of two shapes using quasi-conformal distortion and\ndifferences in mean and Gaussian curvatures, thereby providing a natural way\nfor shape classification. Experiments on Platyrrhine molars demonstrate the\neffectiveness of our method and shed light on the interplay between function\nand shape in nature.\n", " A robot that can carry out a natural-language instruction has been a dream\nsince before the Jetsons cartoon series imagined a life of leisure mediated by\na fleet of attentive robot helpers. It is a dream that remains stubbornly\ndistant. However, recent advances in vision and language methods have made\nincredible progress in closely related areas. This is significant because a\nrobot interpreting a natural-language navigation instruction on the basis of\nwhat it sees is carrying out a vision and language process that is similar to\nVisual Question Answering. Both tasks can be interpreted as visually grounded\nsequence-to-sequence translation problems, and many of the same methods are\napplicable. To enable and encourage the application of vision and language\nmethods to the problem of interpreting visually-grounded navigation\ninstructions, we present the Matterport3D Simulator -- a large-scale\nreinforcement learning environment based on real imagery. Using this simulator,\nwhich can in future support a range of embodied vision and language tasks, we\nprovide the first benchmark dataset for visually-grounded natural language\nnavigation in real buildings -- the Room-to-Room (R2R) dataset.\n", " We investigate algorithmic control of a large swarm of mobile particles (such\nas robots, sensors, or building material) that move in a 2D workspace using a\nglobal input signal (such as gravity or a magnetic field). We show that a maze\nof obstacles to the environment can be used to create complex systems. We\nprovide a wide range of results for a wide range of questions. These can be\nsubdivided into external algorithmic problems, in which particle configurations\nserve as input for computations that are performed elsewhere, and internal\nlogic problems, in which the particle configurations themselves are used for\ncarrying out computations. For external algorithms, we give both negative and\npositive results. If we are given a set of stationary obstacles, we prove that\nit is NP-hard to decide whether a given initial configuration of unit-sized\nparticles can be transformed into a desired target configuration. Moreover, we\nshow that finding a control sequence of minimum length is PSPACE-complete. We\nalso work on the inverse problem, providing constructive algorithms to design\nworkspaces that efficiently implement arbitrary permutations between different\nconfigurations. For internal logic, we investigate how arbitrary computations\ncan be implemented. We demonstrate how to encode dual-rail logic to build a\nuniversal logic gate that concurrently evaluates and, nand, nor, and or\noperations. Using many of these gates and appropriate interconnects, we can\nevaluate any logical expression. However, we establish that simulating the full\nrange of complex interactions present in arbitrary digital circuits encounters\na fundamental difficulty: a fan-out gate cannot be generated. We resolve this\nmissing component with the help of 2x1 particles, which can create fan-out\ngates that produce multiple copies of the inputs. Using these gates we provide\nrules for replicating arbitrary digital circuits.\n", " Training an agent to solve control tasks directly from high-dimensional\nimages with model-free reinforcement learning (RL) has proven difficult. A\npromising approach is to learn a latent representation together with the\ncontrol policy. However, fitting a high-capacity encoder using a scarce reward\nsignal is sample inefficient and leads to poor performance. Prior work has\nshown that auxiliary losses, such as image reconstruction, can aid efficient\nrepresentation learning. However, incorporating reconstruction loss into an\noff-policy learning algorithm often leads to training instability. We explore\nthe underlying reasons and identify variational autoencoders, used by previous\ninvestigations, as the cause of the divergence. Following these findings, we\npropose effective techniques to improve training stability. This results in a\nsimple approach capable of matching state-of-the-art model-free and model-based\nalgorithms on MuJoCo control tasks. Furthermore, our approach demonstrates\nrobustness to observational noise, surpassing existing approaches in this\nsetting. Code, results, and videos are anonymously available at\nhttps://sites.google.com/view/sac-ae/home.\n", " Model-free reinforcement learning has recently been shown to successfully\nlearn navigation policies from raw sensor data. In this work, we address the\nproblem of learning driving policies for an autonomous agent in a high-fidelity\nsimulator. Building upon recent research that applies deep reinforcement\nlearning to navigation problems, we present a modular deep reinforcement\nlearning approach to predict the steering angle of the car from raw images. The\nfirst module extracts a low-dimensional latent semantic representation of the\nimage. The control module trained with reinforcement learning takes the latent\nvector as input to predict the correct steering angle. The experimental results\nhave showed that our method is capable of learning to maneuver the car without\nany human control signals.\n", " One of the focus areas of modern scientific research is to reveal mysteries\nrelated to genes and their interactions. The dynamic interactions between genes\ncan be encoded into a gene regulatory network (GRN), which can be used to gain\nunderstanding on the genetic mechanisms behind observable phenotypes. GRN\ninference from time series data has recently been a focus area of systems\nbiology. Due to low sampling frequency of the data, this is a notoriously\ndifficult problem. We tackle the challenge by introducing the so-called\ncontinuous-time Gaussian process dynamical model, based on Gaussian process\nframework that has gained popularity in nonlinear regression problems arising\nin machine learning. The model dynamics are governed by a stochastic\ndifferential equation, where the dynamics function is modelled as a Gaussian\nprocess. We prove the existence and uniqueness of solutions of the stochastic\ndifferential equation. We derive the probability distribution for the Euler\ndiscretised trajectories and establish the convergence of the discretisation.\nWe develop a GRN inference method called BINGO, based on the developed\nframework. BINGO is based on MCMC sampling of trajectories of the GPDM and\nestimating the hyperparameters of the covariance function of the Gaussian\nprocess. Using benchmark data examples, we show that BINGO is superior in\ndealing with poor time resolution and it is computationally feasible.\n", " Unique existence of analytically strong solutions to stochastic partial\ndifferential equations (SPDE) with drift given by the subdifferential of a\nquasi-convex function and with general multiplicative noise is proven. The\nproof applies a genuinely new method of weighted Galerkin approximations based\non the \"distance\" defined by the quasi-convex function. Spatial regularization\nof the initial condition analogous to the deterministic case is obtained. The\nresults yield a unified framework which is applied to stochastic generalized\nporous media equations, stochastic generalized reaction diffusion equations and\nstochastic generalized degenerated p-Laplace equations. In particular, higher\nregularity for solutions of such SPDE is obtained.\n", " We present a real-time algorithm that finds the Penetration Depth (PD)\nbetween general polygonal models based on iterative and local optimization\ntechniques. Given an in-collision configuration of an object in configuration\nspace, we find an initial collision-free configuration using several methods\nsuch as centroid difference, maximally clear configuration, motion coherence,\nrandom configuration, and sampling-based search. We project this configuration\non to a local contact space using a variant of continuous collision detection\nalgorithm and construct a linear convex cone around the projected\nconfiguration. We then formulate a new projection of the in-collision\nconfiguration onto the convex cone as a Linear Complementarity Problem (LCP),\nwhich we solve using a type of Gauss-Seidel iterative algorithm. We repeat this\nprocedure until a locally optimal PD is obtained. Our algorithm can process\ncomplicated models consisting of tens of thousands triangles at interactive\nrates.\n" ]

[ " Neural codes serve as a language for neurons in the brain. Convex codes,\nwhich arise from the pattern of intersections of convex sets in Euclidean\nspace, are of particular relevance to neuroscience. Not every code is convex,\nhowever, and the combinatorial properties of a code that determine its\nconvexity are still poorly understood. Here we find that a code that can be\nrealized by a collection of open convex sets may or may not be realizable by\nclosed convex sets, and vice versa, establishing that open convex and closed\nconvex codes are distinct classes. We also prove that max intersection-complete\ncodes (i.e. codes that contain all intersections of maximal codewords) are both\nopen convex and closed convex, and provide an upper bound for their minimal\nembedding dimension. Finally, we show that the addition of non-maximal\ncodewords to an open convex code preserves convexity.\n", " Grinberg defined Nyldon words as those words which cannot be factorized into\na sequence of lexicographically nondecreasing smaller Nyldon words. He was\ninspired by Lyndon words, defined the same way except with \"nondecreasing\"\nreplaced by \"nonincreasing.\" Charlier, Philibert, and Stipulanti proved that,\nlike Lyndon words, any word has a unique nondecreasing factorization into\nNyldon words. They also show that the Nyldon words form a right Lazard set, and\nequivalently, a right Hall set. In this paper, we provide a new proof of unique\nfactorization into Nyldon words related to Hall set theory and resolve several\nquestions of Charlier et al. In particular, we prove that Nyldon words of a\nfixed length form a circular code, we prove a result on factorizing powers of\nwords into Nyldon words, and we investigate the Lazard procedure for generating\nNyldon words. We show that these results generalize to a new class of Hall\nsets, of which Nyldon words are an example, that we name \"Nyldon-like sets.\"\n", " We consider a mobile edge computing scenario where users want to perform a\nlinear inference operation $\\boldsymbol{W} \\boldsymbol{x}$ on local data\n$\\boldsymbol{x}$ for some network-side matrix $\\boldsymbol{W}$. The inference\nis performed in a distributed fashion over multiple servers at the network\nedge. For this scenario, we propose a coding scheme that combines a rateless\ncode to provide resiliency against straggling servers--hence reducing the\ncomputation latency--and an irregular-repetition code to provide spatial\ndiversity--hence reducing the communication latency. We further derive a lower\nbound on the total latency--comprising computation latency, communication\nlatency, and decoding latency. The proposed scheme performs remarkably close to\nthe bound and yields significantly lower latency than the scheme based on\nmaximum distance separable codes recently proposed by Zhang and Simeone.\n", " In this paper, we model nested polar code construction as a Markov decision\nprocess (MDP), and tackle it with advanced reinforcement learning (RL)\ntechniques. First, an MDP environment with state, action, and reward is defined\nin the context of polar coding. Specifically, a state represents the\nconstruction of an $(N,K)$ polar code, an action specifies its reduction to an\n$(N,K-1)$ subcode, and reward is the decoding performance. A neural network\narchitecture consisting of both policy and value networks is proposed to\ngenerate actions based on the observed states, aiming at maximizing the overall\nrewards. A loss function is defined to trade off between exploitation and\nexploration. To further improve learning efficiency and quality, an `integrated\nlearning' paradigm is proposed. It first employs a genetic algorithm to\ngenerate a population of (sub-)optimal polar codes for each $(N,K)$, and then\nuses them as prior knowledge to refine the policy in RL. Such a paradigm is\nshown to accelerate the training process, and converge at better performances.\nSimulation results show that the proposed learning-based polar constructions\nachieve comparable, or even better, performances than the state of the art\nunder successive cancellation list (SCL) decoders. Last but not least, this is\nachieved without exploiting any expert knowledge from polar coding theory in\nthe learning algorithms.\n", " We present simulation results on the structure and dynamics of micromagnetic\npoint singularities with atomistic resolution. This is achieved by embedding an\natomistic computational region into a standard micromagnetic algorithm. Several\nlength scales are bridged by means of an adaptive mesh refinement and a\nseamless coupling between the continuum theory and a Heisenberg formulation for\nthe atomistic region. The code operates on graphical processing units and is\nable to detect and track the position of strongly inhomogeneous magnetic\nregions. This enables us to reliably simulate the dynamics of Bloch points,\nwhich means that a fundamental class of micromagnetic switching processes can\nbe analyzed with unprecedented accuracy. We test the code by comparing it with\nestablished results and present its functionality with the example of a\nsimulated field-driven Bloch point motion in a soft-magnetic cylinder.\n", " Background: We study the sparsification of dynamic programming folding\nalgorithms of RNA structures. Sparsification applies to the mfe-folding of RNA\nstructures and can lead to a significant reduction of time complexity. Results:\nWe analyze the sparsification of a particular decomposition rule, $\\Lambda^*$,\nthat splits an interval for RNA secondary and pseudoknot structures of fixed\ntopological genus. Essential for quantifying the sparsification is the size of\nits so called candidate set. We present a combinatorial framework which allows\nby means of probabilities of irreducible substructures to obtain the expected\nsize of the set of $\\Lambda^*$-candidates. We compute these expectations for\narc-based energy models via energy-filtered generating functions (GF) for RNA\nsecondary structures as well as RNA pseudoknot structures. For RNA secondary\nstructures we also consider a simplified loop-energy model. This combinatorial\nanalysis is then compared to the expected number of $\\Lambda^*$-candidates\nobtained from folding mfe-structures. In case of the mfe-folding of RNA\nsecondary structures with a simplified loop energy model our results imply that\nsparsification provides a reduction of time complexity by a constant factor of\n91% (theory) versus a 96% reduction (experiment). For the \"full\" loop-energy\nmodel there is a reduction of 98% (experiment).\n", " To enumerate 3-manifold triangulations with a given property, one typically\nbegins with a set of potential face pairing graphs (also known as dual\n1-skeletons), and then attempts to flesh each graph out into full\ntriangulations using an exponential-time enumeration. However, asymptotically\nmost graphs do not result in any 3-manifold triangulation, which leads to\nsignificant \"wasted time\" in topological enumeration algorithms. Here we give a\nnew algorithm to determine whether a given face pairing graph supports any\n3-manifold triangulation, and show this to be fixed parameter tractable in the\ntreewidth of the graph.\n We extend this result to a \"meta-theorem\" by defining a broad class of\nproperties of triangulations, each with a corresponding fixed parameter\ntractable existence algorithm. We explicitly implement this algorithm in the\nmost generic setting, and we identify heuristics that in practice are seen to\nmitigate the large constants that so often occur in parameterised complexity,\nhighlighting the practicality of our techniques.\n", " Phase retrieval in real or complex Hilbert spaces is the task of recovering a\nvector, up to an overall unimodular multiplicative constant, from magnitudes of\nlinear measurements. In this paper, we assume that the vector is normalized,\nbut retain only qualitative, binary information about the measured magnitudes\nby comparing them with a threshold. In more specific, geometric terms, we\nchoose a sequence of subspaces in a real or complex Hilbert space and only\nrecord whether a given vector is closer to the subspace than to the\ncomplementary subspace. The subspaces have half the dimension of the Hilbert\nspace and are independent, uniformly distributed with respect to the action of\nthe orthogonal or unitary groups. The main goal of this paper is to find a\nfeasible algorithm for approximate recovery based on the information gained\nabout the vector from these binary questions and to establish error bounds for\nits approximate recovery. We provide a pointwise bound for fixed input vectors\nand a uniform bound that controls the worst-case scenario among all inputs.\nBoth bounds hold with high probability with respect to the choice of the\nsubspaces. For real or complex vectors of dimension $n$, the pointwise bound\nrequires $m \\ge C \\delta^{-2} n \\log(n)$ and the uniform bound $m \\ge C\n\\delta^{-2} n^2 \\log(\\delta^{-1} n)$ binary questions in order to achieve an\naccuracy of $\\delta$. The accuracy $\\delta$ is measured by the operator norm of\nthe difference between the rank-one orthogonal projections corresponding to the\nnormalized input vector and its approximate recovery.\n", " We propose a new approach to model composition, based on reducing several\nmodels to the same level of complexity and subsequent combining them together.\nFirstly, we suggest a set of model reduction tools that can be systematically\napplied to a given model. Secondly, we suggest a notion of a minimal complexity\nmodel. This model is the simplest one that can be obtained from the original\nmodel using these tools and still able to approximate experimental data.\nThirdly, we propose a strategy for composing the reduced models together.\nConnection with the detailed model is preserved, which can be advantageous in\nsome applications. A toolbox for model reduction and composition has been\nimplemented as part of the BioUML software and tested on the example of\nintegrating two previously published models of the CD95 (APO-1/Fas) signaling\npathways. We show that the reduced models lead to the same dynamical behavior\nof observable species and the same predictions as in the precursor models. The\ncomposite model is able to recapitulate several experimental datasets which\nwere used by the authors of the original models to calibrate them separately,\nbut also has new dynamical properties.\n", " In this article the reduction of a $n$-dimensional space to a $k$-dimensional\nspace is considered as a reduction of $N^n$ states to $N^k$ states, where $N$\nstands for the number of single-particle states per unit of spatial length. It\nturns out, this space reduction could be understood as another definition of\ninflation. It is shown that the introduction of the non-associativity of the\nalgebra of physical fields in a homogeneous space leads to a nonlinear\nequation, the solutions of which can be considered as two-stage inflation.\nUsing the example of reduction $T\\times R^7$ to $T\\times R^3$, it is shown that\nthere is a continuous cross-linking of the Friedmann and inflationary stages of\nalgebraic inflation at times $10^{-15}$ with the number of baryons $10^{80}$ in\nthe Universe. In this paper, we construct a new gravitational constant based on\na nonassociative octonion algebra.\n", " We introduce a novel method to determine the orientation heterogeneity (mean\ntilt angle and orientational distribution) of molecules at interfaces using\nheterodyne two-dimensional sum frequency generation spectroscopy. By doing so,\nwe not only have solved the long-standing \"magic angle\" challenge, i.e. the\nmeasurement of molecular orientation by assuming a narrow orientational\ndistribution results in ambiguities, but we also are able to determine the\norientational distribution, which is otherwise difficult to measure. We applied\nour new method to a CO2 reduction catalyst/gold interface and found that the\ncatalysts formed a monolayer with a mean tilt angle between the quasi-C3\nsymmetric axis of the catalysts and the surface normal of 53 deg, with 5 deg\norientational distribution. Although applied to a specific system, this method\nis a general way to determine the orientation heterogeneity of an\nensemble-averaged molecular interface, which can potentially be applied to a\nwide-range of energy material, catalytic and biological interfaces.\n", " The switch chain is a well-known Markov chain for sampling directed graphs\nwith a given degree sequence. While not ergodic in general, we show that it is\nergodic for regular degree sequences. We then prove that the switch chain is\nrapidly mixing for regular directed graphs of degree d, where d is any positive\ninteger-valued function of the number of vertices. We bound the mixing time by\nbounding the eigenvalues of the chain. A new result is presented and applied to\nbound the smallest (most negative) eigenvalue. This result is a modification of\na lemma by Diaconis and Stroock, and by using it we avoid working with a lazy\nchain. A multicommodity flow argument is used to bound the second-largest\neigenvalue of the chain. This argument is based on the analysis of a related\nMarkov chain for undirected regular graphs by Cooper, Dyer and Greenhill, but\nwith significant extension required.\n", " We give a self-contained proof of the $A_2$ conjecture, which claims that the\nnorm of any Calderon-Zygmund operator is bounded by the first degree of the\n$A_2$ norm of the weight. The original proof of this result by the first author\nrelied on a subtle and rather difficult reduction to a testing condition by the\nlast three authors. Here we replace this reduction by a new weighted norm bound\nfor dyadic shifts - linear in the $A_2$ norm of the weight and quadratic in the\ncomplexity of the shift -, which is based on a new quantitative two-weight\ninequality for the shifts. These sharp one- and two-weight bounds for dyadic\nshifts are the main new results of this paper. They are obtained by rethinking\nthe corresponding previous results of Lacey-Petermichl-Reguera and\nNazarov-Treil-Volberg. To complete the proof of the $A_2$ conjecture, we also\nprovide a simple variant of the representation, already in the original proof,\nof an arbitrary Calderon-Zygmund operator as an average of random dyadic shifts\nand random dyadic paraproducts. This method of the representation amounts to\nthe refinement of the techniques from nonhomogeneous Harmonic Analysis.\n", " The Separating Hyperplane theorem is a fundamental result in Convex Geometry\nwith myriad applications. Our first result, Random Separating Hyperplane\nTheorem (RSH), is a strengthening of this for polytopes. $\\rsh$ asserts that if\nthe distance between $a$ and a polytope $K$ with $k$ vertices and unit diameter\nin $\\Re^d$ is at least $\\delta$, where $\\delta$ is a fixed constant in $(0,1)$,\nthen a randomly chosen hyperplane separates $a$ and $K$ with probability at\nleast $1/poly(k)$ and margin at least $\\Omega \\left(\\delta/\\sqrt{d} \\right)$.\nAn immediate consequence of our result is the first near optimal bound on the\nerror increase in the reduction from a Separation oracle to an Optimization\noracle over a polytope.\n RSH has algorithmic applications in learning polytopes. We consider a\nfundamental problem, denoted the ``Hausdorff problem'', of learning a unit\ndiameter polytope $K$ within Hausdorff distance $\\delta$, given an optimization\noracle for $K$. Using RSH, we show that with polynomially many random queries\nto the optimization oracle, $K$ can be approximated within error $O(\\delta)$.\nTo our knowledge this is the first provable algorithm for the Hausdorff\nProblem. Building on this result, we show that if the vertices of $K$ are\nwell-separated, then an optimization oracle can be used to generate a list of\npoints, each within Hausdorff distance $O(\\delta)$ of $K$, with the property\nthat the list contains a point close to each vertex of $K$. Further, we show\nhow to prune this list to generate a (unique) approximation to each vertex of\nthe polytope. We prove that in many latent variable settings, e.g., topic\nmodeling, LDA, optimization oracles do exist provided we project to a suitable\nSVD subspace. Thus, our work yields the first efficient algorithm for finding\napproximations to the vertices of the latent polytope under the\nwell-separatedness assumption.\n", " We employ a propagator technique to derive a new relativistic $1/\\qq$\nexpansion of the structure function of a nucleus, composed of point-nucleons.\nWe exploit non-relativistic features of low-momentum nucleons in the target and\nonly treat relativistically the nucleon after absorption of a high-momentum\nvirtual photon. The new series permits a 3-dimensional reduction of each term\nand a formal summation of all Final State Interaction terms. We then show that\na relativistic structure function can be obtained from its non-relativistic\nanalog by a mere change of a scaling variable and an addition of an energy\nshift. We compare the obtained result with an ad hoc generalized\nGersch-Rodriguez-Smith theory, previously used in computations of nuclear\nstructure functions.\n", " A 1965 problem due to Danzer asks whether there exists a set in Euclidean\nspace with finite density intersecting any convex body of volume one. A recent\napproach to this problem is concerned with the construction of dense forests\nand is obtained by a suitable weakening of the volume constraint. A dense\nforest is a discrete point set of finite density getting uniformly close to\nlong enough line segments. The distribution of points in a dense forest is then\nquantified in terms of a visibility function. Another way to weaken the\nassumptions in Danzer's problem is by relaxing the density constraint. In this\nrespect, a new concept is introduced in this paper, namely that of an optical\nforest. An optical forest in $\\mathbb{R}^{d}$ is a point set with optimal\nvisibility but not necessarily with finite density. In the literature, the best\nconstructions of Danzer sets and dense forests lack effectivity. The goal of\nthis paper is to provide constructions of dense and optical forests which yield\nthe best known results in any dimension $d \\ge 2$ both in terms of visibility\nand density bounds and effectiveness. Namely, there are three main results in\nthis work: (1) the construction of a dense forest with the best known\nvisibility bound which, furthermore, enjoys the property of being\ndeterministic; (2) the deterministic construction of an optical forest with a\ndensity failing to be finite only up to a logarithm and (3) the construction of\na planar Peres-type forest (that is, a dense forest obtained from a\nconstruction due to Peres) with the best known visibility bound. This is\nachieved by constructing a deterministic digital sequence satisfying strong\ndispersion properties.\n", " Question answering (QA) over knowledge bases (KBs) is challenging because of\nthe diverse, essentially unbounded, types of reasoning patterns needed.\nHowever, we hypothesize in a large KB, reasoning patterns required to answer a\nquery type reoccur for various entities in their respective subgraph\nneighborhoods. Leveraging this structural similarity between local\nneighborhoods of different subgraphs, we introduce a semiparametric model\n(CBR-SUBG) with (i) a nonparametric component that for each query, dynamically\nretrieves other similar $k$-nearest neighbor (KNN) training queries along with\nquery-specific subgraphs and (ii) a parametric component that is trained to\nidentify the (latent) reasoning patterns from the subgraphs of KNN queries and\nthen apply them to the subgraph of the target query. We also propose an\nadaptive subgraph collection strategy to select a query-specific compact\nsubgraph, allowing us to scale to full Freebase KB containing billions of\nfacts. We show that CBR-SUBG can answer queries requiring subgraph reasoning\npatterns and performs competitively with the best models on several KBQA\nbenchmarks. Our subgraph collection strategy also produces more compact\nsubgraphs (e.g. 55\\% reduction in size for WebQSP while increasing answer\nrecall by 4.85\\%)\\footnote{Code, model, and subgraphs are available at\n\\url{https://github.com/rajarshd/CBR-SUBG}}.\n", " Deep reinforcement learning agents are notoriously sample inefficient, which\nconsiderably limits their application to real-world problems. Recently, many\nmodel-based methods have been designed to address this issue, with learning in\nthe imagination of a world model being one of the most prominent approaches.\nHowever, while virtually unlimited interaction with a simulated environment\nsounds appealing, the world model has to be accurate over extended periods of\ntime. Motivated by the success of Transformers in sequence modeling tasks, we\nintroduce IRIS, a data-efficient agent that learns in a world model composed of\na discrete autoencoder and an autoregressive Transformer. With the equivalent\nof only two hours of gameplay in the Atari 100k benchmark, IRIS achieves a mean\nhuman normalized score of 1.046, and outperforms humans on 10 out of 26 games,\nsetting a new state of the art for methods without lookahead search. To foster\nfuture research on Transformers and world models for sample-efficient\nreinforcement learning, we release our code and models at\nhttps://github.com/eloialonso/iris.\n", " Datasets with non-trivial large scale topology can be hard to embed in\nlow-dimensional Euclidean space with existing dimensionality reduction\nalgorithms. We propose to model topologically complex datasets using vector\nbundles, in such a way that the base space accounts for the large scale\ntopology, while the fibers account for the local geometry. This allows one to\nreduce the dimensionality of the fibers, while preserving the large scale\ntopology. We formalize this point of view, and, as an application, we describe\nan algorithm which takes as input a dataset together with an initial\nrepresentation of it in Euclidean space, assumed to recover part of its large\nscale topology, and outputs a new representation that integrates local\nrepresentations, obtained through local linear dimensionality reduction, along\nthe initial global representation. We demonstrate this algorithm on examples\ncoming from dynamical systems and chemistry. In these examples, our algorithm\nis able to learn topologically faithful embeddings of the data in lower target\ndimension than various well known metric-based dimensionality reduction\nalgorithms.\n", " The existence of an extremal self-dual binary linear code C of length 72 is a\nlong-standing open problem. We continue the investigation of its automorphism\ngroup: looking at the combination of the subcodes fixed by different\ninvolutions and doing a computer calculation with Magma, we prove that Aut(C)\nis not isomorphic to the elementary abelian group of order 8. Combining this\nwith the known results in the literature one obtains that Aut(C) has order at\nmost 5.\n", " The existence of a mechanism within the non-Abelian dynamics of QCD that\nconfines quarks and gluons to the interior of hadrons has long been accepted\nempirically. To explore what this mechanism might look like, this paper\nexamines field-strength descriptions for an extended system of SU(2) charges\nwith spherical symmetry and imposes alternate confining boundary conditions to\nthe time-independent Yang-Mills Maxwell equations. Three types of global\nsolutions to the set of equations can be distinguished: types 0,1,and2. Type-0\nsolutions evade the nonlinear dynamics associated with the radial magnetic\nfield to describe a topologically trivial bound state. Type-1 and type-2\nsolutions both require a domain wall of topological charge to separate the\ninterior volume containing the SU(2) charge densities from the exterior volume\nwhere the boundary conditions are imposed. Type-1 solutions describe an\nexterior volume with a radial magnetic field while type-2 solutions contain a\nsterile exterior volume where all field-strengths vanish. These solutions both\ndescribe a non-Abelian \"spherical dual topological insulator\". the dimensional\nreduction associated with the imposition of spherical system of SU(2) charges\ncan also be applied to SU(3) charges so this simple exercise is directly\nrelevant to understanding confinement in QCD.\n", " A new large-scale parallel multiconfigurational self-consistent field (MCSCF)\nimplementation in the open-source NWChem computational chemistry code is\npresented. The generalized active space (GAS) approach is used to partition\nlarge configuration interaction (CI) vectors and generate a sufficient number\nof batches that can be distributed to the available nodes. Massively parallel\nCI calculations with large active spaces can be treated. The performance of the\nnew parallel MCSCF implementation is presented for the chromium trimer and for\nan active space of 20 electrons in 20 orbitals. Unprecedented CI calculations\nwith an active space of 22 electrons in 22 orbitals for the pentacene systems\nwere performed and a single CI iteration calculation with an active space of 24\nelectrons in 24 orbitals for the chromium tetramer was possible. The chromium\ntetramer corresponds to a CI expansion of one trillion SDs (914 058 513 424)\nand is largest conventional CI calculation attempted up to date.\n", " In this paper, the p-generalized modified error function is defined as the\nsolution to a non-linear ordinary differential problem of second order with a\nRobin type condition at x=0. Existence and uniqueness of a non-negative\nC^\\infty solution is proved by using a fixed point strategy. It is shown that\nthe p-generalized modified error function converges to the p-modified error\nfunction defined as the solution to a similar problem with a Dirichlet\ncondition at x=0. In both problems, for p=1, the generalized modified error\nfunction and the modified error function, studied recently in literature, are\nrecovered. In addition, existence and uniqueness of solution to a problem with\na Neumann condition is also analysed.\n", " A minimally rigid graph, also called Laman graph, models a planar framework\nwhich is rigid for a general choice of distances between its vertices. In other\nwords, there are finitely many ways, up to isometries, to realize such a graph\nin the plane. Using ideas from algebraic and tropical geometry, we derive a\nrecursive formula for the number of such realizations. Combining computational\nresults with the construction of new rigid graphs via gluing techniques, we can\ngive a new lower bound on the maximal possible number of realizations for\ngraphs with a given number of vertices.\n", " Prior knowledge about the imaging physics provides a mechanistic forward\noperator that plays an important role in image reconstruction, although myriad\nsources of possible errors in the operator could negatively impact the\nreconstruction solutions. In this work, we propose to embed the traditional\nmechanistic forward operator inside a neural function, and focus on modeling\nand correcting its unknown errors in an interpretable manner. This is achieved\nby a conditional generative model that transforms a given mechanistic operator\nwith unknown errors, arising from a latent space of self-organizing clusters of\npotential sources of error generation. Once learned, the generative model can\nbe used in place of a fixed forward operator in any traditional\noptimization-based reconstruction process where, together with the inverse\nsolution, the error in prior mechanistic forward operator can be minimized and\nthe potential source of error uncovered. We apply the presented method to the\nreconstruction of heart electrical potential from body surface potential. In\ncontrolled simulation experiments and in-vivo real data experiments, we\ndemonstrate that the presented method allowed reduction of errors in the\nphysics-based forward operator and thereby delivered inverse reconstruction of\nheart-surface potential with increased accuracy.\n", " We propose a new kind of sliding-block puzzle, called Gourds, where the\nobjective is to rearrange 1 x 2 pieces on a hexagonal grid board of 2n + 1\ncells with n pieces, using sliding, turning and pivoting moves. This puzzle has\na single empty cell on a board and forms a natural extension of the 15-puzzle\nto include rotational moves. We analyze the puzzle and completely characterize\nthe cases when the puzzle can always be solved. We also study the complexity of\ndetermining whether a given set of colored pieces can be placed on a colored\nhexagonal grid board with matching colors. We show this problem is NP-complete\nfor arbitrarily many colors, but solvable in randomized polynomial time if the\nnumber of colors is a fixed constant.\n", " The program FIESTA has been completely rewritten. Now it can be used not only\nas a tool to evaluate Feynman integrals numerically, but also to expand Feynman\nintegrals automatically in limits of momenta and masses with the use of sector\ndecompositions and Mellin-Barnes representations. Other important improvements\nto the code are complete parallelization (even to multiple computers),\nhigh-precision arithmetics (allowing to calculate integrals which were undoable\nbefore), new integrators and Speer sectors as a strategy, the possibility to\nevaluate more general parametric integrals.\n", " Reinforcement learning (RL) has shown great promise in optimizing long-term\nuser interest in recommender systems. However, existing RL-based recommendation\nmethods need a large number of interactions for each user to learn a robust\nrecommendation policy. The challenge becomes more critical when recommending to\nnew users who have a limited number of interactions. To that end, in this\npaper, we address the cold-start challenge in the RL-based recommender systems\nby proposing a meta-level model-based reinforcement learning approach for fast\nuser adaptation. In our approach, we learn to infer each user's preference with\na user context variable that enables recommendation systems to better adapt to\nnew users with few interactions. To improve adaptation efficiency, we learn to\nrecover the user policy and reward from only a few interactions via an inverse\nreinforcement learning method to assist a meta-level recommendation agent.\nMoreover, we model the interaction relationship between the user model and\nrecommendation agent from an information-theoretic perspective. Empirical\nresults show the effectiveness of the proposed method when adapting to new\nusers with only a single interaction sequence. We further provide a theoretical\nanalysis of the recommendation performance bound.\n", " This paper studies the properties of a new lower bound for the natural\npseudo-distance. The natural pseudo-distance is a dissimilarity measure between\nshapes, where a shape is viewed as a topological space endowed with a\nreal-valued continuous function. Measuring dissimilarity amounts to minimizing\nthe change in the functions due to the application of homeomorphisms between\ntopological spaces, with respect to the $L_\\infty$-norm. In order to obtain the\nlower bound, a suitable metric between size functions, called matching\ndistance, is introduced. It compares size functions by solving an optimal\nmatching problem between countable point sets. The matching distance is shown\nto be resistant to perturbations, implying that it is always smaller than the\nnatural pseudo-distance. We also prove that the lower bound so obtained is\nsharp and cannot be improved by any other distance between size functions.\n", " In the 1970s and 1980s, searches performed by L. Carter, C. Lam, L. Thiel,\nand S. Swiercz showed that projective planes of order ten with weight 16\ncodewords do not exist. These searches required highly specialized and\noptimized computer programs and required about 2,000 hours of computing time on\nmainframe and supermini computers. In 2011, these searches were verified by D.\nRoy using an optimized C program and 16,000 hours on a cluster of desktop\nmachines. We performed a verification of these searches by reducing the problem\nto the Boolean satisfiability problem (SAT). Our verification uses the\ncube-and-conquer SAT solving paradigm, symmetry breaking techniques using the\ncomputer algebra system Maple, and a result of Carter that there are ten\nnonisomorphic cases to check. Our searches completed in about 30 hours on a\ndesktop machine and produced nonexistence proofs of about 1 terabyte in the\nDRAT (deletion resolution asymmetric tautology) format.\n", " This paper is concerned with the utilization of deterministically modeled\nchemical reaction networks for the implementation of (feed-forward) neural\nnetworks. We develop a general mathematical framework and prove that the\nordinary differential equations (ODEs) associated with certain reaction network\nimplementations of neural networks have desirable properties including (i)\nexistence of unique positive fixed points that are smooth in the parameters of\nthe model (necessary for gradient descent), and (ii) fast convergence to the\nfixed point regardless of initial condition (necessary for efficient\nimplementation). We do so by first making a connection between neural networks\nand fixed points for systems of ODEs, and then by constructing reaction\nnetworks with the correct associated set of ODEs. We demonstrate the theory by\nconstructing a reaction network that implements a neural network with a\nsmoothed ReLU activation function, though we also demonstrate how to generalize\nthe construction to allow for other activation functions (each with the\ndesirable properties listed previously). As there are multiple types of\n\"networks\" utilized in this paper, we also give a careful introduction to both\nreaction networks and neural networks, in order to disambiguate the overlapping\nvocabulary in the two settings and to clearly highlight the role of each\nnetwork's properties.\n", " Establishing the CCZ-equivalence of a pair of APN functions is generally\nquite difficult. In some cases, when seeking to show that a putative new\ninfinite family of APN functions is CCZ inequivalent to an already known\nfamily, we rely on computer calculation for small values of n. In this paper we\npresent a method to prove the inequivalence of quadratic APN functions with the\nGold functions. Our main result is that a quadratic function is CCZ-equivalent\nto an APN Gold function if and only if it is EA-equivalent to that Gold\nfunction. As an application of this result, we prove that a trinomial family of\nAPN functions that exist on finite fields of order 2^n where n = 2 mod 4 are\nCCZ inequivalent to the Gold functions. The proof relies on some knowledge of\nthe automorphism group of a code associated with such a function.\n", " In this work, we obtain a Lyapunov-type and a Hartman-Wintner-type\ninequalities for a linear and a nonlinear fractional differential equation with\ngeneralized Hilfer operator subject to Dirichlet-type boundary conditions. We\nprove existence of positive solutions to a nonlinear fractional boundary value\nproblem. As an application, we obtain a lower bound for the eigenvalues of\ncorresponding equations.\n", " Iterative deepening search is used in applications where the best cost bound\nfor state-space search is unknown. The iterative deepening process is used to\navoid overshooting the appropriate cost bound and doing too much work as a\nresult. However, iterative deepening search also does too much work if the cost\nbound grows too slowly. This paper proposes a new framework for iterative\ndeepening search called exponential-binary state-space search. The approach\ninterleaves exponential and binary searches to find the desired cost bound,\nreducing the worst-case overhead from polynomial to logarithmic.\nExponential-binary search can be used with bounded depth-first search to\nimprove the worst-case performance of IDA* and with breadth-first heuristic\nsearch to improve the worst-case performance of search with inconsistent\nheuristics.\n", " What role can AI play in supporting and constraining creative coding by\nfamilies? To investigate these questions, we built a Wizard of Oz platform to\nhelp families engage in creative coding in partnership with a\nresearcher-operated AI Friend. We designed a 3 week series of programming\nactivities with ten children, 7 to 12 years old, and nine parents. Using a\ncreative self efficacy lens, we observe that families found it easier to\ngenerate game ideas when prompted with questions by AI Friend; parents played a\nunique role in guiding children in more complex programming tasks when the AI\nFriend failed to help, and children were more encouraged to write code for\nnovel ideas using the AI friend help. These findings suggest that AI supported\nplatforms should highlight unique family AI interactions focused on children's\nagency and creative self-efficacy.\n", " We give an effective uniform bound on the multigraded regularity of a\nsubscheme of a smooth projective toric variety X with a given multigraded\nHilbert polynomial. To establish this bound, we introduce a new combinatorial\ntool, called a Stanley filtration, for studying monomial ideals in the\nhomogeneous coordinate ring of X. As a special case, we obtain a new proof of\nGotzmann's regularity theorem. We also discuss applications of this bound to\nthe construction of multigraded Hilbert schemes.\n", " Topological photonics sheds light on some of the surprising phenomena seen in\ncondensed matter physics that arise with the appearance of topological\ninvariants. Optical waveguides provide a well-controlled platform to\ninvestigate effects that relate to different topological phases of matter,\nproviding insight into phenomena such as topological insulators and\nsuperconductors by direct simulation of the states that are protected by the\ntopology of the system. Here, we observe a mode associated with a topological\ndefect in the bulk of a 2D photonic material by introducing a vortex distortion\nto an hexagonal lattice and analogous to graphene. These observations are made\npossible by advances in our experimental methods. We were able to manufacture\nuniform large two-dimensional photonic crystal structures, containing thousands\nof identical waveguides arranged in two dimensions, and we developed a new\nmethod to excite multiples of these waveguides with a well-defined light field.\nThis allows us to probe the detailed spatial features of topological defect\nmodes for the first time. The observed modes lie mid-gap at zero energy and are\nclosely related to Majorana bound states in superconducting vortices. This is\nthe first experimental demonstration of a mode that is a solution to the Dirac\nequation in the presence of a vortex, as proposed by Jackiw and Rossi.\n", " We show that c-planarity is solvable in quadratic time for flat clustered\ngraphs with three clusters if the combinatorial embedding of the underlying\ngraph is fixed. In simpler graph-theoretical terms our result can be viewed as\nfollows. Given a graph $G$ with the vertex set partitioned into three parts\nembedded on a 2-sphere, our algorithm decides if we can augment $G$ by adding\nedges without creating an edge-crossing so that in the resulting spherical\ngraph the vertices of each part induce a connected sub-graph. We proceed by a\nreduction to the problem of testing the existence of a perfect matching in\nplanar bipartite graphs. We formulate our result in a slightly more general\nsetting of cyclic clustered graphs, i.e., the simple graph obtained by\ncontracting each cluster, where we disregard loops and multi-edges, is a cycle.\n", " We define strict confluent drawing, a form of confluent drawing in which the\nexistence of an edge is indicated by the presence of a smooth path through a\nsystem of arcs and junctions (without crossings), and in which such a path, if\nit exists, must be unique. We prove that it is NP-complete to determine whether\na given graph has a strict confluent drawing but polynomial to determine\nwhether it has an outerplanar strict confluent drawing with a fixed vertex\nordering (a drawing within a disk, with the vertices placed in a given order on\nthe boundary).\n", " In this paper we study the state complexity of catenation combined with\nsymmetric difference. First, an upper bound is computed using some combinatoric\ntools. Then, this bound is shown to be tight by giving a witness for it.\nMoreover, we relate this work with the study of state complexity for two other\ncombinations: catenation with union and catenation with intersection. And we\nextract a unified approach which allows to obtain the state complexity of any\ncombination involving catenation and a binary boolean operation.\n", " In the realm of deep learning, the self-attention mechanism has substantiated\nits pivotal role across a myriad of tasks, encompassing natural language\nprocessing and computer vision. Despite achieving success across diverse\napplications, the traditional self-attention mechanism primarily leverages\nlinear transformations for the computation of query, key, and value (QKV),\nwhich may not invariably be the optimal choice under specific circumstances.\nThis paper probes into a novel methodology for QKV computation-implementing a\nspecially-designed neural network structure for the calculation. Utilizing a\nmodified Marian model, we conducted experiments on the IWSLT 2017\nGerman-English translation task dataset and juxtaposed our method with the\nconventional approach. The experimental results unveil a significant\nenhancement in BLEU scores with our method. Furthermore, our approach also\nmanifested superiority when training the Roberta model with the Wikitext-103\ndataset, reflecting a notable reduction in model perplexity compared to its\noriginal counterpart. These experimental outcomes not only validate the\nefficacy of our method but also reveal the immense potential in optimizing the\nself-attention mechanism through neural network-based QKV computation, paving\nthe way for future research and practical applications. The source code and\nimplementation details for our proposed method can be accessed at\nhttps://github.com/ocislyjrti/NeuralAttention.\n", " Motivated by automated junction recognition in tracking data, we study a\nproblem of placing a square or disc of fixed size in an arrangement of lines or\nline segments in the plane. We let distances among the intersection points of\nthe lines and line segments with the square or circle define a clustering, and\nstudy the complexity of \\emph{critical} placements for this clustering. Here\ncritical means that arbitrarily small movements of the placement change the\nclustering.\n A parameter $\\varepsilon$ defines the granularity of the clustering. Without\nany assumptions on $\\varepsilon$, the critical placements have a trivial\n$O(n^4)$ upper bound. When the square or circle has unit size and $0 <\n\\varepsilon < 1$ is given, we show a refined $O(n^2/\\varepsilon^2)$ bound,\nwhich is tight in the worst case.\n We use our combinatorial bounds to design efficient algorithms to compute\njunctions. As a proof of concept for our algorithms we have a prototype\nimplementation that showcases their application in a basic visualization of a\nset of $n$ trajectories and their $k$ most important junctions.\n", " A result of Foata and Schutzenberger states that two statistics on\npermutations, the number of inversions and the inverse major index, have the\nsame distribution on a descent class. We give a multivariate generalization of\nthis property: the sorted vectors of the Lehmer code, of the inverse majcode,\nand of a new code (the inverse saillance code), have the same distribution on a\ndescent class, and their common multivariate generating function is a flagged\nribbon Schur function.\n", " In this paper we are proving the following fact. Let P be an arbitrary simple\npolygon, and let S be an arbitrary set of 15 points inside P. Then there exists\na subset T of S that is not \"visually discernible\", that is, T is not equal to\nthe intersection of S with the visibility region vis(v) of any point v in P. In\nother words, the VC-dimension d of visibility regions in a simple polygon\ncannot exceed 14. Since Valtr proved in 1998 that d \\in [6,23] holds, no\nprogress has been made on this bound. By epsilon-net theorems our reduction\nimmediately implies a smaller upper bound to the number of guards needed to\ncover P.\n", " Recently, both the ATLAS and CMS experiments have observed an excess of\nevents that could be the first evidence for a 125 GeV Higgs boson. This is a\nfew GeV below the (absolute) vacuum stability bound on the Higgs mass in the\nStandard Model (SM), assuming a Planck mass ultraviolet (UV) cutoff. In this\npaper, we study some implications of a 125 GeV Higgs boson for new physics in\nterms of the vacuum stability bound. We first consider the seesaw extension of\nthe SM and find that in type III seesaw, the vacuum stability bound on the\nHiggs mass can be as low as 125 GeV for the seesaw scale around a TeV. Next we\ndicuss some alternative new physics models which provide an effective\nultraviolet cutoff lower than the Planck mass. An effective cutoff \\Lambda\n\\simeq 10^{11} GeV leads to a vacuum stability bound on the Higgs mass of 125\nGeV. In a gauge-Higgs unification scenario with five-dimensional flat\nspacetime, the so-called gauge-Higgs condition allows us to predict a Higgs\nmass of 125 GeV, with the compactification scale of the extra-dimension being\nidentified as the cutoff scale \\Lambda \\simeq 10^{11} GeV. Identifying the\ncompactification scale with the unification scale of the SM SU(2) gauge\ncoupling and the top quark Yukawa coupling yields a Higgs mass of 121\\pm 2 GeV.\n", " The paper describes two general problems encountered in computational\nassignments at the introductory level. First, novice students often treat\ncomputer code as almost magic incantations, and like novices in many fields,\nhave trouble creating new algorithms or procedures to solve novel problems.\nSecond, the nature of computational studies often means that the results\ngenerated are interpreted via theoretically devised quantities, which may not\nmeet a student's internal standards for proof when compared to an experimental\nmeasurement.\n The paper then offers a lab/programming assignment, used in a calculus-based\nphysics course, which was devised to address these problems. In the assignment,\nstudents created a computational model of the heat flow involved in heating an\niron rod in a blacksmith's forge. After creating the simulation, students\nattended a blacksmithing seminar and had a chance to work with iron and take\ndata on its heating in a coke-fueled forge. On their return to campus, students\nrevised their computational models in light of their experimental data.\n", " A record in a permutation is a maximum or a minimum, from the left or from\nthe right. The entries of a permutation can be partitioned into two types: the\nones that are records are called external points, the others are called\ninternal points. Permutations without internal points have been studied under\nthe name of square permutations. Here, we explore permutations with a fixed\nnumber of internals points, called almost square permutations. Unlike with\nsquare permutations, a precise enumeration for the total number of almost\nsquare permutations of size $n+k$ with exactly $k$ internal points is not\nknown. However, using a probabilistic approach, we are able to determine the\nasymptotic enumeration. This allows us to describe the permuton limit of almost\nsquare permutations with $k$ internal points, both when $k$ is fixed and when\n$k$ tends to infinity along a negligible sequence with respect to the size of\nthe permutation. Finally, we show that our techniques are quite general by\nstudying the set of $321$-avoiding permutations of size $n$ with exactly $k$\nadditional internal points ($k$ fixed). In this case we obtain an interesting\nasymptotic enumeration in terms of the Brownian excursion area. As a\nconsequence, we show that the points of a uniform permutation in this set\nconcentrate on the diagonal and the fluctuations of these points converge in\ndistribution to a biased Brownian excursion.\n", " We characterize inclusions of compact noncommutative convex sets with the\nproperty that every continuous affine function on the smaller set can be\nextended to a continuous affine function on the larger set with a uniform\nbound. As an application of this result, we obtain a simple geometric\ncharacterization of (possibly nonunital) operator systems that are dualizable,\nmeaning that their dual can be equipped with an operator system structure. We\nfurther establish some permanence properties of dualizability, and provide a\nlarge new class of dualizable operator systems. These results are new even when\nspecialized to ordinary compact convex sets.\n", " SARAH is a Mathematica package optimized for the fast, efficient and precise\nstudy of supersymmetric models beyond the MSSM: a new model can be defined in a\nshort form and all vertices are derived. This allows SARAH to create model\nfiles for FeynArts/FormCalc, CalcHep/CompHep and WHIZARD/OMEGA. The newest\nversion of SARAH now provides the possibility to create model files in the UFO\nformat which is supported by MadGraph 5, MadAnalysis, GoSam, and soon by\nHerwig++. Furthermore, SARAH also calculates the mass matrices, RGEs and\none-loop corrections to the mass spectrum. This information is used to write\nsource code for SPheno in order to create a precision spectrum generator for\nthe given model. This spectrum-generator-generator functionality as well as the\noutput of WHIZARD and CalcHep model files have seen further improvement in this\nversion. Also models including Dirac Gauginos are supported with the new\nversion of SARAH, and additional checks for the consistency of model\nimplementations have been created.\n", " Given a digraph $G$, a lot of attention has been deserved on the maximum\nnumber $\\phi(G)$ of fixed points in a Boolean network $f:\\{0,1\\}^n\\to\\{0,1\\}^n$\nwith $G$ as interaction graph. In particular, a central problem in network\ncoding consists in studying the optimality of the classical upper bound\n$\\phi(G)\\leq 2^{\\tau}$, where $\\tau$ is the minimum size of a feedback vertex\nset of $G$. In this paper, we study the maximum number $\\phi_m(G)$ of fixed\npoints in a {\\em monotone} Boolean network with interaction graph $G$. We\nestablish new upper and lower bounds on $\\phi_m(G)$ that depends on the cycle\nstructure of $G$. In addition to $\\tau$, the involved parameters are the\nmaximum number $\\nu$ of vertex-disjoint cycles, and the maximum number\n$\\nu^{*}$ of vertex-disjoint cycles verifying some additional technical\nconditions. We improve the classical upper bound $2^\\tau$ by proving that\n$\\phi_m(G)$ is at most the largest sub-lattice of $\\{0,1\\}^\\tau$ without chain\nof size $\\nu+1$, and without another forbidden-pattern of size $2\\nu^{*}$.\nThen, we prove two optimal lower bounds: $\\phi_m(G)\\geq \\nu+1$ and\n$\\phi_m(G)\\geq 2^{\\nu^{*}}$. As a consequence, we get the following\ncharacterization: $\\phi_m(G)=2^\\tau$ if and only if $\\nu^{*}=\\tau$. As another\nconsequence, we get that if $c$ is the maximum length of a chordless cycle of\n$G$ then $2^{\\nu/3^c}\\leq\\phi_m(G)\\leq 2^{c\\nu}$. Finally, with the technics\nintroduced, we establish an upper bound on the number of fixed points of any\nBoolean network according to its signed interaction graph.\n" ]

[ " Recent interest in integrated sensing and communications has led to the\ndesign of novel signal processing techniques to recover information from an\noverlaid radar-communications signal. Here, we focus on a spectral coexistence\nscenario, wherein the channels and transmit signals of both radar and\ncommunications systems are unknown to the common receiver. In this dual-blind\ndeconvolution (DBD) problem, the receiver admits a multi-carrier wireless\ncommunications signal that is overlaid with the radar signal reflected off\nmultiple targets. The communications and radar channels are represented by\ncontinuous-valued range-times or delays corresponding to multiple transmission\npaths and targets, respectively. Prior works addressed recovery of unknown\nchannels and signals in this ill-posed DBD problem through atomic norm\nminimization but contingent on individual minimum separation conditions for\nradar and communications channels. In this paper, we provide an optimal joint\nseparation condition using extremal functions from the Beurling-Selberg\ninterpolation theory. Thereafter, we formulate DBD as a low-rank modified\nHankel matrix retrieval and solve it via nuclear norm minimization. We estimate\nthe unknown target and communications parameters from the recovered low-rank\nmatrix using multiple signal classification (MUSIC) method. We show that the\njoint separation condition also guarantees that the underlying Vandermonde\nmatrix for MUSIC is well-conditioned. Numerical experiments validate our\ntheoretical findings.\n", " We study nuclear effects of charged current deep inelastic neutrino-iron\nscattering in the framework of a chi^2 analysis of parton distribution\nfunctions (PDFs). We extract a set of iron PDFs which are used to compute\nx_Bj-dependent and Q^2-dependent nuclear correction factors for iron structure\nfunctions which are required in global analyses of free nucleon PDFs. We\ncompare our results with nuclear correction factors from neutrino-nucleus\nscattering models and correction factors for charged lepton-iron scattering. We\nfind that, except for very high x_Bj, our correction factors differ in both\nshape and magnitude from the correction factors of the models and\ncharged-lepton scattering.\n", " We propose a method that morphs high-orger meshes such that their boundaries\nand interfaces coincide/align with implicitly defined geometries. Our focus is\nparticularly on the case when the target surface is prescribed as the zero\nisocontour of a smooth discrete function. Common examples of this scenario\ninclude using level set functions to represent material interfaces in\nmultimaterial configurations, and evolving geometries in shape and topology\noptimization. The proposed method formulates the mesh optimization problem as a\nvariational minimization of the sum of a chosen mesh-quality metric using the\nTarget-Matrix Optimization Paradigm (TMOP) and a penalty term that weakly\nforces the selected faces of the mesh to align with the target surface. The\ndistinct features of the method are use of a source mesh to represent the level\nset function with sufficient accuracy, and adaptive strategies for setting the\npenalization weight and selecting the faces of the mesh to be fit to the target\nisocontour of the level set field. We demonstrate that the proposed method is\nrobust for generating boundary- and interface-fitted meshes for curvilinear\ndomains using different element types in 2D and 3D.\n", " We report on a dramatic improvement of the performance of the classical\ntime-delayed autosynchronization method (TDAS) to control unstable steady\nstates, by applying a time-varying delay in the TDAS control scheme in a form\nof a deterministic or stochastic delay-modulation in a fixed interval around a\nnominal value $T_0$. The successfulness of this variable delay feedback control\n(VDFC) is illustrated by a numerical control simulation of the Lorenz and\nR\\\"{o}ssler systems using three different types of time-delay modulations: a\nsawtooth wave, a sine wave, and a uniform random distribution. We perform a\ncomparative analysis between the VDFC method and the standard TDAS method for a\nsawtooth-wave modulation by analytically determining the domains of control for\nthe generic case of an unstable fixed point of focus type.\n", " The nuclear modification factor $R_{AA}$ for $\\pi_0$ production in Au+Au\ncollisions at $\\sqrt{s}=200$ AGeV is calculated, and studied at high transverse\nmomenta $p_T$. The soft thermalized nuclear medium is described within the\nframework of relativistic ideal three-dimensional hydrodynamics. The energy\nloss of partonic jets is evaluated in the context of gluon bremsstrahlung in\nthe thermalized partonic matter. We provide a systematic analysis of the\nazimuthal asymmetry of $\\pi_0$ suppression at high $p_T$ in central and\nnon-central collisions, at mid and forward rapidity. The determination of\n$R_{AA}$ as a function of $p_T$, at different azimuthal angles, and different\nrapidities makes for a stringent test of our theoretical understanding of jet\nenergy loss over a variety of in-medium path lengths, temperatures and initial\npartonic jet energies. This lays the groundwork for a precise tomography of the\nnuclear medium.\n", " In this study, a hybrid method based on an Adaptive Neuro-Fuzzy Inference\nSystem (ANFIS) and Particle Swarm Optimization (PSO) for diagnosing Liver\ndisorders (ANFIS-PSO) is introduced. This smart diagnosis method deals with a\ncombination of making an inference system and optimization process which tries\nto tune the hyper-parameters of ANFIS based on the data-set. The Liver diseases\ncharacteristics are taken from the UCI Repository of Machine Learning\nDatabases. The number of these characteristic attributes are 7, and the sample\nnumber is 354. The right diagnosis performance of the ANFIS-PSO intelligent\nmedical system for liver disease is evaluated by using classification accuracy,\nsensitivity and specificity analysis, respectively. According to the\nexperimental results, the performance of ANFIS-PSO can be more considerable\nthan traditional FIS and ANFIS without optimization phase.\n", " This paper addresses the problem of statistical inference for latent\ncontinuous-time stochastic processes, which is often intractable, particularly\nfor discrete state space processes described by Markov jump processes. To\novercome this issue, we propose a new tractable inference scheme based on an\nentropic matching framework that can be embedded into the well-known\nexpectation propagation algorithm. We demonstrate the effectiveness of our\nmethod by providing closed-form results for a simple family of approximate\ndistributions and apply it to the general class of chemical reaction networks,\nwhich are a crucial tool for modeling in systems biology. Moreover, we derive\nclosed form expressions for point estimation of the underlying parameters using\nan approximate expectation maximization procedure. We evaluate the performance\nof our method on various chemical reaction network instantiations, including a\nstochastic Lotka-Voltera example, and discuss its limitations and potential for\nfuture improvements. Our proposed approach provides a promising direction for\naddressing complex continuous-time Bayesian inference problems.\n", " Few-shot learning has become essential for producing models that generalize\nfrom few examples. In this work, we identify that metric scaling and metric\ntask conditioning are important to improve the performance of few-shot\nalgorithms. Our analysis reveals that simple metric scaling completely changes\nthe nature of few-shot algorithm parameter updates. Metric scaling provides\nimprovements up to 14% in accuracy for certain metrics on the mini-Imagenet\n5-way 5-shot classification task. We further propose a simple and effective way\nof conditioning a learner on the task sample set, resulting in learning a\ntask-dependent metric space. Moreover, we propose and empirically test a\npractical end-to-end optimization procedure based on auxiliary task co-training\nto learn a task-dependent metric space. The resulting few-shot learning model\nbased on the task-dependent scaled metric achieves state of the art on\nmini-Imagenet. We confirm these results on another few-shot dataset that we\nintroduce in this paper based on CIFAR100. Our code is publicly available at\nhttps://github.com/ElementAI/TADAM.\n", " We develop an efficient algorithm, coordinate descent FCI (CDFCI), for the\nelectronic structure ground state calculation in the configuration interaction\nframework. CDFCI solves an unconstrained non-convex optimization problem, which\nis a reformulation of the full configuration interaction eigenvalue problem,\nvia an adaptive coordinate descent method with a deterministic compression\nstrategy. CDFCI captures and updates appreciative determinants with different\nfrequencies proportional to their importance. We show that CDFCI produces\naccurate variational energy for both static and dynamic correlation by\nbenchmarking the binding curve of nitrogen dimer in the cc-pVDZ basis with\n$10^{-3}$ mHa accuracy. We also demonstrate the efficiency and accuracy of\nCDFCI for strongly correlated chromium dimer in the Ahlrichs VDZ basis and\nproduces state-of-the-art variational energy.\n", " A method of analyzing time bounds for randomized distributed algorithms is\npresented, in the context of a new and general framework for describing and\nreasoning about randomized algorithms. The method consists of proving auxiliary\nstatements of the form U (t)->(p) U', which means that whenever the algorithm\nbegins in a state in set U, with probability p, it will reach a state in set U'\nwithin time t. The power of the method is illustrated by its use in proving a\nconstant upper bound on the expected time for some process to reach its\ncritical region, in Lehmann and Rabin's Dining Philosophers algorithm.\n", " This article considers the problem of solving a system of $n$ real polynomial\nequations in $n+1$ variables. We propose an algorithm based on Newton's method\nand subdivision for this problem. Our algorithm is intended only for\nnondegenerate cases, in which case the solution is a 1-dimensional curve. Our\nfirst main contribution is a definition of a condition number measuring\nreciprocal distance to degeneracy that can distinguish poor and well\nconditioned instances of this problem. (Degenerate problems would be infinitely\nill conditioned in our framework.) Our second contribution, which is the main\nnovelty of our algorithm, is an analysis showing that its running time is\nbounded in terms of the condition number of the problem instance as well as $n$\nand the polynomial degrees.\n", " The concept of microstructure of the real space, considered as a mathematical\nlattice of cells (the tessel-lattice), and notions of canonical particles and\nfields, which are generated by the space, are analyzed. Submicroscopic\nmechanics based on this concept is discussed and employed for in-depth study of\nthe nucleon-nucleon interaction. It is argued that a deformation coat is\ndeveloped in the real space around the nucleon (as is the case with any other\ncanonical particle such as electron, muon, etc.) and that they are the\ndeformation coats that are responsible for the appearance of nuclear forces.\nOne more source of nuclear forces is associated with inerton clouds,\nexcitations of the space tessel-lattice (the excitations are a substructure of\nnucleons' matter waves), which accompany moving nucleons as in the case of any\nother canonical particles. Two nuclear systems are under consideration: the\ndeuteron and a weight nucleus. It is shown that a weight nucleus is a cluster\nof interacting protons and neutrons. The condition of the cluster stability is\nobtained in the framework of the statistical mechanical approach. On the whole,\nthe paper proposes a radically new approach to a very old unsolved problem, the\norigin of the nuclear forces, allowing the understanding mechanism(s) occurring\nat low energy nuclear transmutations.\n", " We present a new method based on functional tensor decomposition and dynamic\ntensor approximation to compute the solution of a high-dimensional\ntime-dependent nonlinear partial differential equation (PDE). The idea of\ndynamic approximation is to project the time derivative of the PDE solution\nonto the tangent space of a low-rank functional tensor manifold at each time.\nSuch a projection can be computed by minimizing a convex energy functional over\nthe tangent space. This minimization problem yields the unique optimal velocity\nvector that allows us to integrate the PDE forward in time on a tensor manifold\nof constant rank. In the case of initial/boundary value problems defined in\nreal separable Hilbert spaces, this procedure yields evolution equations for\nthe tensor modes in the form of a coupled system of one-dimensional\ntime-dependent PDEs. We apply the dynamic tensor approximation to a\nfour-dimensional Fokker-Planck equation with non-constant drift and diffusion\ncoefficients, and demonstrate its accuracy in predicting relaxation to\nstatistical equilibrium.\n", " High quality upsampling of sparse 3D point clouds is critically useful for a\nwide range of geometric operations such as reconstruction, rendering, meshing,\nand analysis. In this paper, we propose a data-driven algorithm that enables an\nupsampling of 3D point clouds without the need for hard-coded rules. Our\napproach uses a deep network with Chamfer distance as the loss function,\ncapable of learning the latent features in point clouds belonging to different\nobject categories. We evaluate our algorithm across different amplification\nfactors, with upsampling learned and performed on objects belonging to the same\ncategory as well as different categories. We also explore the desirable\ncharacteristics of input point clouds as a function of the distribution of the\npoint samples. Finally, we demonstrate the performance of our algorithm in\nsingle-category training versus multi-category training scenarios. The final\nproposed model is compared against a baseline, optimization-based upsampling\nmethod. Results indicate that our algorithm is capable of generating more\nuniform and accurate upsamplings.\n", " We give a self-contained proof of the $A_2$ conjecture, which claims that the\nnorm of any Calderon-Zygmund operator is bounded by the first degree of the\n$A_2$ norm of the weight. The original proof of this result by the first author\nrelied on a subtle and rather difficult reduction to a testing condition by the\nlast three authors. Here we replace this reduction by a new weighted norm bound\nfor dyadic shifts - linear in the $A_2$ norm of the weight and quadratic in the\ncomplexity of the shift -, which is based on a new quantitative two-weight\ninequality for the shifts. These sharp one- and two-weight bounds for dyadic\nshifts are the main new results of this paper. They are obtained by rethinking\nthe corresponding previous results of Lacey-Petermichl-Reguera and\nNazarov-Treil-Volberg. To complete the proof of the $A_2$ conjecture, we also\nprovide a simple variant of the representation, already in the original proof,\nof an arbitrary Calderon-Zygmund operator as an average of random dyadic shifts\nand random dyadic paraproducts. This method of the representation amounts to\nthe refinement of the techniques from nonhomogeneous Harmonic Analysis.\n", " In this paper we develop a general theory of compressed sensing for analog\nsignals, in close similarity to prior results for vectors in finite dimensional\nspaces that are sparse in a given orthonormal basis. The signals are modeled by\nfunctions in a reproducing kernel Hilbert space. Sparsity is defined as the\nminimal number of terms in expansions based on the kernel functions. Minimizing\nthis number is under certain conditions equivalent to minimizing an atomic\nnorm, the pre-dual of the supremum norm for functions in the Hilbert space. The\nnorm minimizer is shown to exist based on a compactness argument. Recovery\nbased on minimizing the atomic norm is robust and stable, so it provides\ncontrollable accuracy for recovery when the signal is only approximately sparse\nand the measurement is corrupted by noise.\n As applications of the theory, we include results on the recovery of sparse\nbandlimited functions and functions that have a sparse inverse short-time\nFourier transform.\n", " Frames have established themselves as a means to derive redundant, yet stable\ndecompositions of a signal for analysis or transmission, while also promoting\nsparse expansions. However, when the signal dimension is large, the computation\nof the frame measurements of a signal typically requires a large number of\nadditions and multiplications, and this makes a frame decomposition intractable\nin applications with limited computing budget. To address this problem, in this\npaper, we focus on frames in finite-dimensional Hilbert spaces and introduce\nsparsity for such frames as a new paradigm. In our terminology, a sparse frame\nis a frame whose elements have a sparse representation in an orthonormal basis,\nthereby enabling low-complexity frame decompositions. To introduce a precise\nmeaning of optimality, we take the sum of the numbers of vectors needed of this\northonormal basis when expanding each frame vector as sparsity measure. We then\nanalyze the recently introduced algorithm Spectral Tetris for construction of\nunit norm tight frames and prove that the tight frames generated by this\nalgorithm are in fact optimally sparse with respect to the standard unit vector\nbasis. Finally, we show that even the generalization of Spectral Tetris for the\nconstruction of unit norm frames associated with a given frame operator\nproduces optimally sparse frames.\n", " Finding high-quality solutions to mixed-integer linear programming problems\n(MILPs) is of great importance for many practical applications. In this\nrespect, the refinement heuristic local branching (LB) has been proposed to\nproduce improving solutions and has been highly influential for the development\nof local search methods in MILP. The algorithm iteratively explores a sequence\nof solution neighborhoods defined by the so-called local branching constraint,\nnamely, a linear inequality limiting the distance from a reference solution.\nFor a LB algorithm, the choice of the neighborhood size is critical to\nperformance. In this work, we study the relation between the size of the search\nneighborhood and the behavior of the underlying LB algorithm, and we devise a\nleaning based framework for predicting the best size for the specific instance\nto be solved. Furthermore, we have also investigated the relation between the\ntime limit for exploring the LB neighborhood and the actual performance of LB\nscheme, and devised a strategy for adapting the time limit. We computationally\nshow that the neighborhood size and time limit can indeed be learned, leading\nto improved performances and that the overall algorithm generalizes well both\nwith respect to the instance size and, remarkably, across instances.\n", " Online learning algorithms have become a ubiquitous tool in the machine\nlearning toolbox and are frequently used in small, resource-constraint\nenvironments. Among the most successful online learning methods are Decision\nTree (DT) ensembles. DT ensembles provide excellent performance while adapting\nto changes in the data, but they are not resource efficient. Incremental tree\nlearners keep adding new nodes to the tree but never remove old ones increasing\nthe memory consumption over time. Gradient-based tree learning, on the other\nhand, requires the computation of gradients over the entire tree which is\ncostly for even moderately sized trees. In this paper, we propose a novel\nmemory-efficient online classification ensemble called shrub ensembles for\nresource-constraint systems. Our algorithm trains small to medium-sized\ndecision trees on small windows and uses stochastic proximal gradient descent\nto learn the ensemble weights of these `shrubs'. We provide a theoretical\nanalysis of our algorithm and include an extensive discussion on the behavior\nof our approach in the online setting. In a series of 2~959 experiments on 12\ndifferent datasets, we compare our method against 8 state-of-the-art methods.\nOur Shrub Ensembles retain an excellent performance even when only little\nmemory is available. We show that SE offers a better accuracy-memory trade-off\nin 7 of 12 cases, while having a statistically significant better performance\nthan most other methods. Our implementation is available under\nhttps://github.com/sbuschjaeger/se-online .\n", " In this paper, we first define a difference measure between the old and new\nsequential patterns of stream data, which is proved to be a distance. Then we\npropose an experimental method, called TPD (Tradeoff between Performance and\nDifference), to decide when to update the sequential patterns of stream data by\nmaking a tradeoff between the performance of increasingly updating algorithms\nand the difference of sequential patterns. The experiments for the incremental\nupdating algorithm IUS on two data sets show that generally, as the size of\nincremental windows grows, the values of the speedup and the values of the\ndifference will decrease and increase respectively. It is also shown\nexperimentally that the incremental ratio determined by the TPD method does not\nmonotonically increase or decrease but changes in a range between 20 and 30\npercentage for the IUS algorithm.\n", " We present a general and modular algorithmic framework for path planning of\nrobots. Our framework combines geometric methods for exact and complete\nanalysis of low-dimensional configuration spaces, together with practical,\nconsiderably simpler sampling-based approaches that are appropriate for higher\ndimensions. In order to facilitate the transfer of advanced geometric\nalgorithms into practical use, we suggest taking samples that are entire\nlow-dimensional manifolds of the configuration space that capture the\nconnectivity of the configuration space much better than isolated point\nsamples. Geometric algorithms for analysis of low-dimensional manifolds then\nprovide powerful primitive operations. The modular design of the framework\nenables independent optimization of each modular component. Indeed, we have\ndeveloped, implemented and optimized a primitive operation for complete and\nexact combinatorial analysis of a certain set of manifolds, using arrangements\nof curves of rational functions and concepts of generic programming. This in\nturn enabled us to implement our framework for the concrete case of a polygonal\nrobot translating and rotating amidst polygonal obstacles. We demonstrate that\nthe integration of several carefully engineered components leads to significant\nspeedup over the popular PRM sampling-based algorithm, which represents the\nmore simplistic approach that is prevalent in practice. We foresee possible\nextensions of our framework to solving high-dimensional problems beyond motion\nplanning.\n", " The purpose of this seminar, which was presented at the Universitat\nPolit\\`ecnica de Val\\`encia in late 2012, is to explain several results\nconcerning the bounded approximation property for Fr\\'echet spaces. We give a\nfull detailed proof of an important result due to Pe\\l czy\\'nski that asserts\nthat every separable Fr\\'echet space with the bounded approximation property is\nisomorphic to a complemented subspace of a Fr\\'echet space with a Schauder\nbasis. We also explain Vogt's example of a nuclear Fr\\'echet space without the\nbounded approximation property. This example is simpler than the original\ncounterexample due to Dubinski. These examples solved a long standing problem\nof Grothendieck. Vogt obtained later another simple example of a nuclear\nFr\\'echet function space without the bounded approximation property. The\nrelation of the bounded approximation property for Fr\\'echet spaces with a\ncontinuous norm and the countably normable spaces, including several results\ndue to Dubinski and Vogt, is also explained.\n", " Motivation: Laboratory gene regulatory data for a species are sporadic.\nDespite the abundance of gene regulatory network algorithms that employ single\ndata sets, few algorithms can combine the vast but disperse sources of data and\nextract the potential information. With a motivation to compensate for this\nshortage, we developed an algorithm called GENEREF that can accumulate\ninformation from multiple types of data sets in an iterative manner, with each\niteration boosting the performance of the prediction results.\n Results: The algorithm is examined extensively on data extracted from the\nquintuple DREAM4 networks and DREAM5's Escherichia coli and Saccharomyces\ncerevisiae networks and sub-networks. Many single-dataset and multi-dataset\nalgorithms were compared to test the performance of the algorithm. Results show\nthat GENEREF surpasses non-ensemble state-of-the-art multi-perturbation\nalgorithms on the selected networks and is competitive to present\nmultiple-dataset algorithms. Specifically, it outperforms dynGENIE3 and is on\npar with iRafNet. Also, we argued that a scoring method solely based on the\nAUPR criterion would be more trustworthy than the traditional score.\n Availability: The Python implementation along with the data sets and results\ncan be downloaded from github.com/msaremi/GENEREF\n", " Clarksons algorithm is a two-staged randomized algorithm for solving linear\nprograms. This algorithm has been simplified and adapted to fit the framework\nof LP-type problems. In this framework we can tackle a number of non-linear\nproblems such as computing the smallest enclosing ball of a set of points in\nR^d . In 2006, it has been shown that the algorithm in its original form works\nfor violator spaces too, which are a proper general- ization of LP-type\nproblems. It was not clear, however, whether previous simplifications of the\nalgorithm carry over to the new setting. In this paper we show the following\ntheoretical results: (a) It is shown, for the first time, that Clarksons second\nstage can be simplified. (b) The previous simplifications of Clarksons first\nstage carry over to the violator space setting. (c) Furthermore, we show the\nequivalence of violator spaces and partitions of the hypercube by hypercubes.\n", " A major challenge in collaborative filtering methods is how to produce\nrecommendations for cold items (items with no ratings), or integrate cold item\ninto an existing catalog. Over the years, a variety of hybrid recommendation\nmodels have been proposed to address this problem by utilizing items' metadata\nand content along with their ratings or usage patterns. In this work, we wish\nto revisit the cold start problem in order to draw attention to an overlooked\nchallenge: the ability to integrate and balance between (regular) warm items\nand completely cold items. In this case, two different challenges arise: (1)\npreserving high quality performance on warm items, while (2) learning to\npromote cold items to relevant users. First, we show that these two objectives\nare in fact conflicting, and the balance between them depends on the business\nneeds and the application at hand. Next, we propose a novel hybrid\nrecommendation algorithm that bridges these two conflicting objectives and\nenables a harmonized balance between preserving high accuracy for warm items\nwhile effectively promoting completely cold items. We demonstrate the\neffectiveness of the proposed algorithm on movies, apps, and articles\nrecommendations, and provide an empirical analysis of the cold-warm trade-off.\n", " Boolean tensor has been broadly utilized in representing high dimensional\nlogical data collected on spatial, temporal and/or other relational domains.\nBoolean Tensor Decomposition (BTD) factorizes a binary tensor into the Boolean\nsum of multiple rank-1 tensors, which is an NP-hard problem. Existing BTD\nmethods have been limited by their high computational cost, in applications to\nlarge scale or higher order tensors. In this work, we presented a\ncomputationally efficient BTD algorithm, namely \\textit{Geometric Expansion for\nall-order Tensor Factorization} (GETF), that sequentially identifies the rank-1\nbasis components for a tensor from a geometric perspective. We conducted\nrigorous theoretical analysis on the validity as well as algorithemic\nefficiency of GETF in decomposing all-order tensor. Experiments on both\nsynthetic and real-world data demonstrated that GETF has significantly improved\nperformance in reconstruction accuracy, extraction of latent structures and it\nis an order of magnitude faster than other state-of-the-art methods.\n", " Cloud computing as a fairly new commercial paradigm, widely investigated by\ndifferent researchers, already has a great range of challenges. Pricing is a\nmajor problem in Cloud computing marketplace; as providers are competing to\nattract more customers without knowing the pricing policies of each other. To\novercome this lack of knowledge, we model their competition by an\nincomplete-information game. Considering the issue, this work proposes a\npricing policy related to the regret minimization algorithm and applies it to\nthe considered incomplete-information game. Based on the competition based\nmarketplace of the Cloud, providers update the distribution of their strategies\nusing the experienced regret. The idea of iteratively applying the algorithm\nfor updating probabilities of strategies causes the regret get minimized\nfaster. The experimental results show much more increase in profits of the\nproviders in comparison with other pricing policies. Besides, the efficiency of\na variety of regret minimization techniques in a simulated marketplace of Cloud\nare discussed which have not been observed in the studied literature. Moreover,\nreturn on investment of providers in considered organizations is studied and\npromising results appeared.\n", " We present results of testing of the ability of eleven popular scoring\nfunctions to predict native docked positions using a recently developed method\n[1] for estimation the entropy contributions of relative motions to\nprotein-ligand binding affinity. The method is based on the integration of the\nconfigurational integral over clusters obtained from multiple docked positions.\nWe use a test set of 100 PDB protein-ligand complexes and ensembles of 101\ndocked positions generated by Wang et al [2] for each ligand in the test set.\nTo test the suggested method we compare the averaged root-mean square\ndeviations (RMSD) of the top-scored ligand docked positions, accounting and not\naccounting for entropy contributions, relative to the experimentally determined\npositions. We demonstrate that the method increases docking accuracy by 10-21%\nwhen used in conjunction with the AutoDock scoring function, by 2-25% with\nG-Score, by 7-41% with D-Score, by 0-8% with LigScore, by 1-6% with PLP, by\n0-12% with LUDI, by 2-8% with F-Score, by 7-29% with ChemScore, by 0-9% with\nX-Score, by 2-19% with PMF, and by 1-7% with DrugScore. We also compare the\nperformance of the suggested method with the method based on ranking by cluster\noccupancy only. We analyze how the choice of a RMSD-tolerance and a low bound\nof dense clusters impacts on docking accuracy of the scoring methods. We derive\noptimal intervals of the RMSD-tolerance for 11 scoring functions.\n", " Arithmetic complexity has a main role in the performance of algorithms for\nspectrum evaluation. Arithmetic transform theory offers a method for computing\ntrigonometrical transforms with minimal number of multiplications. In this\npaper, the proposed algorithms for the arithmetic Fourier transform are\nsurveyed. A new arithmetic transform for computing the discrete Hartley\ntransform is introduced: the Arithmetic Hartley transform. The interpolation\nprocess is shown to be the key element of the arithmetic transform theory.\n", " Personality is a fundamental construct in psychology, reflecting an\nindividual's behavior, thinking, and emotional patterns. Previous researches\nhave made some progress in personality detection, primarily by utilizing the\nwhole text to predict personality. However, these studies generally tend to\noverlook psychological knowledge: they rarely apply the well-established\ncorrelations between emotion regulation and personality. Based on this, we\npropose a new personality detection method called EERPD. This method introduces\nthe use of emotion regulation, a psychological concept highly correlated with\npersonality, for personality prediction. By combining this feature with emotion\nfeatures, it retrieves few-shot examples and provides process CoTs for\ninferring labels from text. This approach enhances the understanding of LLM for\npersonality within text and improves the performance in personality detection.\nExperimental results demonstrate that EERPD significantly enhances the accuracy\nand robustness of personality detection, outperforming previous SOTA by\n15.05/4.29 in average F1 on the two benchmark datasets.\n", " We propose an efficient deterministic method to calculate the full\nconfiguration interaction (FCI) ground state energy. This method leverages the\nsparseness of the Lanczos basis vectors that span the Krylov subspace\nassociated with the Hamiltonian to rapidly calculate the lowest eigenvalue of\nthe effective Hamiltonian formed in this basis. By iteratively performing the\nspanning and diagonalization steps, this method is capable of rapidly reaching\nchemical accuracy for a variety of strongly correlated molecules, including\nstretched N$_2$ and C$_2$, in merely several tens of iterations. To accelerate\nour algorithm, we implement parallelized sparse matrix-sparse vector\nmultiplication, which features a novel hashing method that establishes a\none-to-one correspondence between the set of full configuration Slater\ndeterminants and a set of consecutive integers. A near-linear acceleration was\nobserved.\n", " Many biological processes discriminate between correct and incorrect\nsubstrates through the kinetic proofreading mechanism which enables lower error\nat the cost of higher energy dissipation. Elucidating physicochemical\nconstraints for global minimization of dissipation and error is important for\nunderstanding enzyme evolution. Here, we identify theoretically a fundamental\nerror-cost bound which tightly constrains the performance of proofreading\nnetworks under any parameter variations preserving the rate discrimination\nbetween substrates. The bound is kinetically controlled, i.e. completely\ndetermined by the difference between the transition state energies on the\nunderlying free energy landscape. The importance of the bound is analyzed for\nthree biological processes. DNA replication by T7 DNA polymerase is shown to be\nnearly optimized, i.e. its kinetic parameters place it in the immediate\nproximity of the error-cost bound. The isoleucyl-tRNA synthetase (IleRS) of E.\ncoli also operates close to the bound, but further optimization is prevented by\nthe need for reaction speed. In contrast, E. coli ribosome operates in a\nhigh-dissipation regime, potentially in order to speed up protein production.\nTogether, these findings establish a fundamental error-dissipation relation in\nbiological proofreading networks and provide a theoretical framework for\nstudying error-dissipation trade-off in other systems with biological\ndiscrimination.\n", " Searching topological similarity between a pair of shapes or data is an\nimportant problem in data analysis and visualization. The problem of computing\nsimilarity measures using scalar topology has been studied extensively and\nproven useful in shape and data matching. Even though multi-field (or\nmultivariate) topology-based techniques reveal richer topological features,\nresearch on computing similarity measures using multi-field topology is still\nin its infancy. In the current paper, we propose a novel similarity measure\nbetween two piecewise-linear multi-fields based on their multi-resolution Reeb\nspaces - a newly developed data-structure that captures the topology of a\nmulti-field. Overall, our method consists of two steps: (i) building a\nmulti-resolution Reeb space corresponding to each of the multi-fields and (ii)\nproposing a similarity measure for a list of matching pairs (of nodes),\nobtained by comparing the multi-resolution Reeb spaces. We demonstrate an\napplication of the proposed similarity measure by detecting the nuclear\nscission point in a time-varying multi-field data from computational physics.\n", " A path or a polygonal domain is C-oriented if the orientations of its edges\nbelong to a set of C given orientations; this is a generalization of the\nnotable rectilinear case (C = 2). We study exact and approximation algorithms\nfor minimum-link C-oriented paths and paths with unrestricted orientations,\nboth in C-oriented and in general domains. Our two main algorithms are as\nfollows:\n A subquadratic-time algorithm with a non-trivial approximation guarantee for\ngeneral (unrestricted-orientation) minimum-link paths in general domains.\n An algorithm to find a minimum-link C-oriented path in a C-oriented domain.\nOur algorithm is simpler and more time-space efficient than the prior\nalgorithm.\n We also obtain several related results:\n - 3SUM-hardness of determining the link distance with unrestricted\norientations (even in a rectilinear domain).\n - An optimal algorithm for finding a minimum-link rectilinear path in a\nrectilinear domain. The algorithm and its analysis are simpler than the\nexisting ones.\n - An extension of our methods to find a C-oriented minimum-link path in a\ngeneral (not necessarily C-oriented) domain.\n - A more efficient algorithm to compute a 2-approximate C-oriented\nminimum-link path.\n - A notion of \"robust\" paths. We show how minimum-link C-oriented paths\napproximate the robust paths with unrestricted orientations to within an\nadditive error of 1.\n", " Recently cloud-based graph convolutional network (GCN) has demonstrated great\nsuccess and potential in many privacy-sensitive applications such as personal\nhealthcare and financial systems. Despite its high inference accuracy and\nperformance on cloud, maintaining data privacy in GCN inference, which is of\nparamount importance to these practical applications, remains largely\nunexplored. In this paper, we take an initial attempt towards this and develop\n$\\textit{CryptoGCN}$--a homomorphic encryption (HE) based GCN inference\nframework. A key to the success of our approach is to reduce the tremendous\ncomputational overhead for HE operations, which can be orders of magnitude\nhigher than its counterparts in the plaintext space. To this end, we develop an\napproach that can effectively take advantage of the sparsity of matrix\noperations in GCN inference to significantly reduce the computational overhead.\nSpecifically, we propose a novel AMA data formatting method and associated\nspatial convolution methods, which can exploit the complex graph structure and\nperform efficient matrix-matrix multiplication in HE computation and thus\ngreatly reduce the HE operations. We also develop a co-optimization framework\nthat can explore the trade offs among the accuracy, security level, and\ncomputational overhead by judicious pruning and polynomial approximation of\nactivation module in GCNs. Based on the NTU-XVIEW skeleton joint dataset, i.e.,\nthe largest dataset evaluated homomorphically by far as we are aware of, our\nexperimental results demonstrate that $\\textit{CryptoGCN}$ outperforms\nstate-of-the-art solutions in terms of the latency and number of homomorphic\noperations, i.e., achieving as much as a 3.10$\\times$ speedup on latency and\nreduces the total Homomorphic Operation Count by 77.4\\% with a small accuracy\nloss of 1-1.5$\\%$.\n", " In this article we consider the problem of testing, for two finite sets of\npoints in the Euclidean space, if their convex hulls are disjoint and computing\nan optimal supporting hyperplane if so. This is a fundamental problem of\nclassification in machine learning known as the hard-margin SVM. The problem\ncan be formulated as a quadratic programming problem. The SMO algorithm is the\ncurrent state of art algorithm for solving it, but it does not answer the\nquestion of separability. An alternative to solving both problems is the\nTriangle Algorithm, a geometrically inspired algorithm, initially described for\nthe convex hull membership problem, a fundamental problem in linear\nprogramming. First, we describe the experimental performance of the Triangle\nAlgorithm for testing the intersection of two convex hulls. Next, we compare\nthe performance of Triangle Algorithm with SMO for finding the optimal\nsupporting hyperplane. Based on experimental results ranging up to 5000 points\nin each set in dimensions up to 10000, the Triangle Algorithm outperforms SMO.\n", " A direct orbital optimization method is presented for density functional\ncalculations of excited electronic states using either a real space grid or a\nplane wave basis set. The method is variational, provides atomic forces in the\nexcited states, and can be applied to Kohn-Sham (KS) functionals as well as\norbital-density dependent functionals (ODD) including explicit self-interaction\ncorrection. The implementation for KS functionals involves two nested loops:\n(1) An inner loop for finding a stationary point in a subspace spanned by the\noccupied and a few virtual orbitals corresponding to the excited state; (2) an\nouter loop for minimizing the energy in a tangential direction in the space of\nthe orbitals. For ODD functionals, a third loop is used to find the unitary\ntransformation that minimizes the energy functional among occupied orbitals\nonly. Combined with the maximum overlap method, the algorithm converges in\nchallenging cases where conventional self-consistent field algorithms tend to\nfail. The benchmark tests presented include two charge-transfer excitations in\nnitrobenzene and an excitation of CO to degenerate $\\pi^\\ast$ orbitals where\nthe importance of complex orbitals is illustrated. An application of the method\nto several metal-to-ligand charge-transfer and metal-centred excited states of\nan Fe$^{\\rm II}$ photosensitizer complex is described and the results compared\nto reported experimental estimates. The method is also used to study the effect\nof Perdew-Zunger self-interaction correction on valence and Rydberg excited\nstates of several molecules, both singlet and triplet states, and the\nperformance compared to semilocal and hybrid functionals.\n", " Recent years have witnessed a large amount of decentralized data in multiple\n(edge) devices of end-users, while the aggregation of the decentralized data\nremains difficult for machine learning jobs due to laws or regulations.\nFederated Learning (FL) emerges as an effective approach to handling\ndecentralized data without sharing the sensitive raw data, while\ncollaboratively training global machine learning models. The servers in FL need\nto select (and schedule) devices during the training process. However, the\nscheduling of devices for multiple jobs with FL remains a critical and open\nproblem. In this paper, we propose a novel multi-job FL framework to enable the\nparallel training process of multiple jobs. The framework consists of a system\nmodel and two scheduling methods. In the system model, we propose a parallel\ntraining process of multiple jobs, and construct a cost model based on the\ntraining time and the data fairness of various devices during the training\nprocess of diverse jobs. We propose a reinforcement learning-based method and a\nBayesian optimization-based method to schedule devices for multiple jobs while\nminimizing the cost. We conduct extensive experimentation with multiple jobs\nand datasets. The experimental results show that our proposed approaches\nsignificantly outperform baseline approaches in terms of training time (up to\n8.67 times faster) and accuracy (up to 44.6% higher).\n", " A key component of automated algorithm selection and configuration, which in\nmost cases are performed using supervised machine learning (ML) methods is a\ngood-performing predictive model. The predictive model uses the feature\nrepresentation of a set of problem instances as input data and predicts the\nalgorithm performance achieved on them. Common machine learning models struggle\nto make predictions for instances with feature representations not covered by\nthe training data, resulting in poor generalization to unseen problems. In this\nstudy, we propose a workflow to estimate the generalizability of a predictive\nmodel for algorithm performance, trained on one benchmark suite to another. The\nworkflow has been tested by training predictive models across benchmark suites\nand the results show that generalizability patterns in the landscape feature\nspace are reflected in the performance space.\n", " We investigate exact crossing minimization for graphs that differ from trees\nby a small number of additional edges, for several variants of the crossing\nminimization problem. In particular, we provide fixed parameter tractable\nalgorithms for the 1-page book crossing number, the 2-page book crossing\nnumber, and the minimum number of crossed edges in 1-page and 2-page book\ndrawings.\n", " Most practitioners use a variant of the Alpha-Beta algorithm, a simple\ndepth-first pro- cedure, for searching minimax trees. SSS*, with its best-first\nsearch strategy, reportedly offers the potential for more efficient search.\nHowever, the complex formulation of the al- gorithm and its alleged excessive\nmemory requirements preclude its use in practice. For two decades, the search\nefficiency of \"smart\" best-first SSS* has cast doubt on the effectiveness of\n\"dumb\" depth-first Alpha-Beta. This paper presents a simple framework for\ncalling Alpha-Beta that allows us to create a variety of algorithms, including\nSSS* and DUAL*. In effect, we formulate a best-first algorithm using\ndepth-first search. Expressed in this framework SSS* is just a special case of\nAlpha-Beta, solving all of the perceived drawbacks of the algorithm. In\npractice, Alpha-Beta variants typically evaluate less nodes than SSS*. A new\ninstance of this framework, MTD(f), out-performs SSS* and NegaScout, the\nAlpha-Beta variant of choice by practitioners.\n", " Stochastic Boolean networks, or more generally, stochastic discrete networks,\nare an important class of computational models for molecular interaction\nnetworks. The stochasticity stems from the updating schedule. Standard updating\nschedules include the synchronous update, where all the nodes are updated at\nthe same time, and the asynchronous update where a random node is updated at\neach time step. The former produces a deterministic dynamics while the latter a\nstochastic dynamics. A more general stochastic setting considers propensity\nparameters for updating each node. Stochastic Discrete Dynamical Systems (SDDS)\nis a modeling framework that considers two propensity parameters for updating\neach node and uses one when the update has a positive impact on the variable,\nthat is, when the update causes the variable to increase its value, and uses\nthe other when the update has a negative impact, that is, when the update\ncauses it to decrease its value. This framework offers additional features for\nsimulations but also adds a complexity in parameter estimation of the\npropensities. This paper presents a method for estimating the propensity\nparameters for SDDS. The method is based on adding noise to the system using\nthe Google PageRank approach to make the system ergodic and thus guaranteeing\nthe existence of a stationary distribution. Then with the use of a genetic\nalgorithm, the propensity parameters are estimated. Approximation techniques\nthat make the search algorithms efficient are also presented and Matlab/Octave\ncode to test the algorithms are available at\nhttp://www.ms.uky.edu/~dmu228/GeneticAlg/Code.html.\n", " We present algorithms for the computation of $\\varepsilon$-coresets for\n$k$-median clustering of point sequences in $\\mathbb{R}^d$ under the\n$p$-dynamic time warping (DTW) distance. Coresets under DTW have not been\ninvestigated before, and the analysis is not directly accessible to existing\nmethods as DTW is not a metric. The three main ingredients that allow our\nconstruction of coresets are the adaptation of the $\\varepsilon$-coreset\nframework of sensitivity sampling, bounds on the VC dimension of approximations\nto the range spaces of balls under DTW, and new approximation algorithms for\nthe $k$-median problem under DTW. We achieve our results by investigating\napproximations of DTW that provide a trade-off between the provided accuracy\nand amenability to known techniques. In particular, we observe that given $n$\ncurves under DTW, one can directly construct a metric that approximates DTW on\nthis set, permitting the use of the wealth of results on metric spaces for\nclustering purposes. The resulting approximations are the first with polynomial\nrunning time and achieve a very similar approximation factor as\nstate-of-the-art techniques. We apply our results to produce a practical\nalgorithm approximating $(k,\\ell)$-median clustering under DTW.\n", " We introduce a tensor-based clustering method to extract sparse,\nlow-dimensional structure from high-dimensional, multi-indexed datasets. This\nframework is designed to enable detection of clusters of data in the presence\nof structural requirements which we encode as algebraic constraints in a linear\nprogram. Our clustering method is general and can be tailored to a variety of\napplications in science and industry. We illustrate our method on a collection\nof experiments measuring the response of genetically diverse breast cancer cell\nlines to an array of ligands. Each experiment consists of a cell line-ligand\ncombination, and contains time-course measurements of the early-signalling\nkinases MAPK and AKT at two different ligand dose levels. By imposing\nappropriate structural constraints and respecting the multi-indexed structure\nof the data, the analysis of clusters can be optimized for biological\ninterpretation and therapeutic understanding. We then perform a systematic,\nlarge-scale exploration of mechanistic models of MAPK-AKT crosstalk for each\ncluster. This analysis allows us to quantify the heterogeneity of breast cancer\ncell subtypes, and leads to hypotheses about the signalling mechanisms that\nmediate the response of the cell lines to ligands.\n", " Gillespie's direct method for stochastic simulation of chemical kinetics is a\nstaple of computational systems biology research. However, the algorithm\nrequires explicit enumeration of all reactions and all chemical species that\nmay arise in the system. In many cases, this is not feasible due to the\ncombinatorial explosion of reactions and species in biological networks.\nRule-based modeling frameworks provide a way to exactly represent networks\ncontaining such combinatorial complexity, and generalizations of Gillespie's\ndirect method have been developed as simulation engines for rule-based modeling\nlanguages. Here, we provide both a high-level description of the algorithms\nunderlying the simulation engines, termed network-free simulation algorithms,\nand how they have been applied in systems biology research. We also define a\ngeneric rule-based modeling framework and describe a number of technical\ndetails required for adapting Gillespie's direct method for network-free\nsimulation. Finally, we briefly discuss potential avenues for advancing\nnetwork-free simulation and the role they continue to play in modeling\ndynamical systems in biology.\n", " We present a model for neutrino-nucleus scattering in the energy region\nrelevant for present and forthcoming neutrino-oscillation experiments. The\nmodel is based on the RPA treatment of the nuclear responses in the\nquasi-elastic and Delta-resonance region. It includes also in a\nphenomenological way nucleon knock-out. It aims at the description, within a\nsingle framework, of several final state channels i.e. quasi-elastic,\nincoherent and coherent one-pion production and two- or several-nucleon\nknock-out.\n", " We analyze the fate of dynamical systems that consist of two kind of\nprocesses. The first type is supposed to perform a certain function by\nprocessing information at a required high accuracy, which is, however, limited\nto less than 100 percent, while the second process serves to maintain the\nrequired precision. Both processes are assumed to be subject to a trade-off\nbetween cost and precision, where the cost have to be paid from renewable but\nlimited resources. In a discrete map we pursue the time evolution of errors and\ndetermine the conditions under which the fate of the system is either a stable\nperformance at the desired accuracy, or a deterioration. Deterioration may be\nrealized either as an accumulation of errors or a decline of resources when\nthey are all absorbed for maintenance. We point to possible implications for\nliving organisms and their perspectives to avoid an accumulation of errors in\nthe course of time.\n", " Large language models (LLMs) have exhibited great potential in mathematical\nreasoning. However, there remains a performance gap in this area between\nexisting open-source models and closed-source models such as GPT-4. In this\npaper, we introduce MathGenie, a novel method for generating diverse and\nreliable math problems from a small-scale problem-solution dataset (denoted as\nseed data). We augment the ground-truth solutions of our seed data and train a\nback-translation model to translate the augmented solutions back into new\nquestions. Subsequently, we generate code-integrated solutions for the new\nquestions. To ensure the correctness of the code-integrated solutions, we\nemploy rationale-based strategy for solution verification. Various pretrained\nmodels, ranging from 7B to 70B, are trained on the newly curated data to test\nthe effectiveness of the proposed augmentation technique, resulting in a family\nof models known as MathGenieLM. These models consistently outperform previous\nopen-source models across five representative mathematical reasoning datasets,\nachieving state-of-the-art performance. In particular, MathGenieLM-InternLM2\nachieves an accuracy of 87.7% on GSM8K and 55.7% on MATH, securing the best\noverall score among open-source language models.\n", " Discrete-state stochastic models have become a well-established approach to\ndescribe biochemical reaction networks that are influenced by the inherent\nrandomness of cellular events. In the last years severalmethods for accurately\napproximating the statistical moments of such models have become very popular\nsince they allow an efficient analysis of complex networks. We propose a\ngeneralized method of moments approach for inferring the parameters of reaction\nnetworks based on a sophisticated matching of the statistical moments of the\ncorresponding stochastic model and the sample moments of population snapshot\ndata. The proposed parameter estimation method exploits recently developed\nmoment-based approximations and provides estimators with desirable statistical\nproperties when a large number of samples is available. We demonstrate the\nusefulness and efficiency of the inference method on two case studies. The\ngeneralized method of moments provides accurate and fast estimations of unknown\nparameters of reaction networks. The accuracy increases when also moments of\norder higher than two are considered. In addition, the variance of the\nestimator decreases, when more samples are given or when higher order moments\nare included.\n", " Recently, utilizing reinforcement learning (RL) to generate molecules with\ndesired properties has been highlighted as a promising strategy for drug\ndesign. A molecular docking program - a physical simulation that estimates\nprotein-small molecule binding affinity - can be an ideal reward scoring\nfunction for RL, as it is a straightforward proxy of the therapeutic potential.\nStill, two imminent challenges exist for this task. First, the models often\nfail to generate chemically realistic and pharmacochemically acceptable\nmolecules. Second, the docking score optimization is a difficult exploration\nproblem that involves many local optima and less smooth surfaces with respect\nto molecular structure. To tackle these challenges, we propose a novel RL\nframework that generates pharmacochemically acceptable molecules with large\ndocking scores. Our method - Fragment-based generative RL with Explorative\nExperience replay for Drug design (FREED) - constrains the generated molecules\nto a realistic and qualified chemical space and effectively explores the space\nto find drugs by coupling our fragment-based generation method and a novel\nerror-prioritized experience replay (PER). We also show that our model performs\nwell on both de novo and scaffold-based schemes. Our model produces molecules\nof higher quality compared to existing methods while achieving state-of-the-art\nperformance on two of three targets in terms of the docking scores of the\ngenerated molecules. We further show with ablation studies that our method,\npredictive error-PER (FREED(PE)), significantly improves the model performance.\n" ]

[ " The gains afforded by cloud radio access network (C-RAN) in terms of savings\nin capital and operating expenses, flexibility, interference management and\nnetwork densification rely on the presence of high-capacity low-latency\nfronthaul connectivity between remote radio heads (RRHs) and baseband unit\n(BBU). In light of the non-uniform and limited availability of fiber optics\ncables, the bandwidth constraints on the fronthaul network call, on the one\nhand, for the development of advanced baseband compression strategies and, on\nthe other hand, for a closer investigation of the optimal functional split\nbetween RRHs and BBU. In this chapter, after a brief introduction to signal\nprocessing challenges in C-RAN, this optimal function split is studied at the\nphysical (PHY) layer as it pertains to two key baseband signal processing\nsteps, namely channel estimation in the uplink and channel encoding/ linear\nprecoding in the downlink. Joint optimization of baseband fronthaul compression\nand of baseband signal processing is tackled under different PHY functional\nsplits, whereby uplink channel estimation and downlink channel encoding/ linear\nprecoding are carried out either at the RRHs or at the BBU. The analysis, based\non information-theoretical arguments, and numerical results yields insight into\nthe configurations of network architecture and fronthaul capacities in which\ndifferent functional splits are advantageous. The treatment also emphasizes the\nversatility of deterministic and stochastic successive convex approximation\nstrategies for the optimization of C-RANs.\n", " A dynamical mean-field theory is developed to analyze stochastic single-cell\ndynamics of gene expression. By explicitly taking account of nonequilibrium and\nnonadiabatic features of the DNA state fluctuation, two-time correlation\nfunctions and response functions of single-cell dynamics are derived. The\nmethod is applied to a self-regulating gene to predict a rich variety of\ndynamical phenomena such as anomalous increase of relaxation time and\noscillatory decay of correlations. Effective \"temperature\" defined as the ratio\nof the correlation to the response in the protein number is small when the DNA\nstate change is frequent, while it grows large when the DNA state change is\ninfrequent, indicating the strong enhancement of noise in the latter case.\n", " Single-cell sequencing has a significant role to explore biological processes\nsuch as embryonic development, cancer evolution, and cell differentiation.\nThese biological properties can be presented by a two-dimensional scatter plot.\nHowever, single-cell sequencing data generally has very high dimensionality.\nTherefore, dimensionality reduction should be used to process the high\ndimensional sequencing data for 2D visualization and subsequent biological\nanalysis. The traditional dimensionality reduction methods, which do not\nconsider the structure characteristics of single-cell sequencing data, are\ndifficult to reveal the data structure in the 2D representation. In this paper,\nwe develop a 2D feature representation method based on graph convolutional\nnetworks (GCN) for the visualization of single-cell data, termed single-cell\nsequencing graph embedding networks (SGEN). This method constructs the graph by\nthe similarity relationship between cells and adopts GCN to analyze the\nneighbor embedding information of samples, which makes the similar cell closer\nto each other on the 2D scatter plot. The results show SGEN achieves obvious 2D\ndistribution and preserves the high-dimensional relationship of different\ncells. Meanwhile, similar cell clusters have spatial continuity rather than\nrelying heavily on random initialization, which can reflect the trajectory of\ncell development in this scatter plot.\n", " We consider the problem of inferring the unknown parameters of a stochastic\nbiochemical network model from a single measured time-course of the\nconcentration of some of the involved species. Such measurements are available,\ne.g., from live-cell fluorescence microscopy in image-based systems biology. In\naddition, fluctuation time-courses from, e.g., fluorescence correlation\nspectroscopy provide additional information about the system dynamics that can\nbe used to more robustly infer parameters than when considering only mean\nconcentrations. Estimating model parameters from a single experimental\ntrajectory enables single-cell measurements and quantification of cell--cell\nvariability. We propose a novel combination of an adaptive Monte Carlo sampler,\ncalled Gaussian Adaptation, and efficient exact stochastic simulation\nalgorithms that allows parameter identification from single stochastic\ntrajectories. We benchmark the proposed method on a linear and a non-linear\nreaction network at steady state and during transient phases. In addition, we\ndemonstrate that the present method also provides an ellipsoidal volume\nestimate of the viable part of parameter space and is able to estimate the\nphysical volume of the compartment in which the observed reactions take place.\n", " In our recent paper we characterized the extension property for symmetrized\nweight composition for linear codes over a module alphabet. Several\nimprovements for the case of vector space alphabets are given in this paper. A\ndetailed description of the property of $G$-pseudo-injectivity for vector\nspaces is made.\n", " Clarksons algorithm is a two-staged randomized algorithm for solving linear\nprograms. This algorithm has been simplified and adapted to fit the framework\nof LP-type problems. In this framework we can tackle a number of non-linear\nproblems such as computing the smallest enclosing ball of a set of points in\nR^d . In 2006, it has been shown that the algorithm in its original form works\nfor violator spaces too, which are a proper general- ization of LP-type\nproblems. It was not clear, however, whether previous simplifications of the\nalgorithm carry over to the new setting. In this paper we show the following\ntheoretical results: (a) It is shown, for the first time, that Clarksons second\nstage can be simplified. (b) The previous simplifications of Clarksons first\nstage carry over to the violator space setting. (c) Furthermore, we show the\nequivalence of violator spaces and partitions of the hypercube by hypercubes.\n", " In the present paper we introduce and study finite point subsets of a special\nkind, called optimum distributions, in the n-dimensional unit cube. Such\ndistributions are closely related with known (delta,s,n)-nets of low\ndiscrepancy. It turns out that optimum distributions have a rich combinatorial\nstructure. Namely, we show that optimum distributions can be characterized\ncompletely as maximum distance separable codes with respect to a non-Hamming\nmetric. Weight spectra of such codes can be evaluated precisely. We also\nconsider linear codes and distributions and study their general properties\nincluding the duality with respect to a suitable inner product. The\ncorresponding generalized MacWilliams identities for weight enumerators are\nbriefly discussed. Broad classes of linear maximum distance separable codes and\nlinear optimum distributions are explicitely constructed in the paper by the\nHermite interpolations over finite fields.\n", " We give a unified approach to handle the problem of extending functions with\nvalues in a locally convex Hausdorff space $E$ over a field $\\mathbb{K}$, which\nhave weak extensions in a space $\\mathcal{F}(\\Omega,\\mathbb{K})$ of\nscalar-valued functions on a set $\\Omega$, to functions in a vector-valued\ncounterpart $\\mathcal{F}(\\Omega,E)$ of $\\mathcal{F}(\\Omega,\\mathbb{K})$. The\nresults obtained base upon a representation of vector-valued functions as\nlinear continuous operators and extend results of Bonet, Frerick, Gramsch and\nJord\\'{a}. In particular, we apply them to obtain a sequence space\nrepresentation of $\\mathcal{F}(\\Omega,E)$ from a known representation of\n$\\mathcal{F}(\\Omega,\\mathbb{K})$.\n", " Globally non-positively curved, or CAT(0), polyhedral complexes arise in a\nnumber of applications, including evolutionary biology and robotics. These\nspaces have unique shortest paths and are composed of Euclidean polyhedra, yet\nmany algorithms and properties of shortest paths and convex hulls in Euclidean\nspace fail to transfer over. We give an algorithm, using linear programming, to\ncompute the convex hull of a set of points in a 2-dimensional CAT(0) polyhedral\ncomplex with a single vertex. We explore the use of shortest path maps to\nanswer single-source shortest path queries in 2-dimensional CAT(0) polyhedral\ncomplexes, and we unify efficient solutions for 2-manifold and rectangular\ncases.\n", " The Tutte polynomial of a connected graph was originally defined by Tutte as\na sum over all spanning trees of monomials depending on a fixed linear order on\nthe set of edges. Tuttle proved that while these monomials do depend on the\nlinear order, the sum does not. The present paper is a result of a reflection\nupon this classical theorem of Tutte. It is devoted to an elementary and\ndetailed proof of this theorem in its natural generality, i.e. not for graphs,\nbut for matroids.\n In contrast with usual methods, the emphasis is on the properties of maps (as\nopposed to elements) naturally associated to a matroid (or a graph) and an\norder. In order to fully explain the four-fold symmetry of the proof, we\nintroduce notion of linking between two matroids on the same set. While every\nmatroid is, in fact, linked only to itself and to its dual matroid, the notion\nof a linking identifies the essential features of the theory and allows to\nreplace the usual four similar arguments by a single one. This is done in\nTheorem 9.1, which is the focal point of the present paper.\n With the exception of Introduction, the present paper is self-contained\nmodulo basic concepts related to sets and maps. In particular, no knowledge of\nthe matroid theory or of the graph theory is assumed.\n", " In 1986, in order to study the linear representations of the braid group\n$B\\_n$coming from the monodromy of the Knizhnik-Zamolodchikov differential\nequations,Drinfel'd introduced a class of formal power series $\\Phi$on\nnoncommutative variables. These formal series can be considered as a class of\nassociators. We here give an interpretation of them as well as some new tools\nover Noncommutative Evolution Equations. Asymptotic phenomena are also\ndiscussed.\n", " We propose a partially learned approach for the solution of ill posed inverse\nproblems with not necessarily linear forward operators. The method builds on\nideas from classical regularization theory and recent advances in deep learning\nto perform learning while making use of prior information about the inverse\nproblem encoded in the forward operator, noise model and a regularizing\nfunctional. The method results in a gradient-like iterative scheme, where the\n\"gradient\" component is learned using a convolutional network that includes the\ngradients of the data discrepancy and regularizer as input in each iteration.\nWe present results of such a partially learned gradient scheme on a non-linear\ntomographic inversion problem with simulated data from both the Sheep-Logan\nphantom as well as a head CT. The outcome is compared against FBP and TV\nreconstruction and the proposed method provides a 5.4 dB PSNR improvement over\nthe TV reconstruction while being significantly faster, giving reconstructions\nof 512 x 512 volumes in about 0.4 seconds using a single GPU.\n", " As one of the newest members in Artificial Immune Systems (AIS), the\nDendritic Cell Algorithm (DCA) has been applied to a range of problems. These\napplications mainly belong to the field of anomaly detection. However,\nreal-time detection, a new challenge to anomaly detection, requires improvement\non the real-time capability of the DCA. To assess such capability, formal\nmethods in the research of rea-time systems can be employed. The findings of\nthe assessment can provide guideline for the future development of the\nalgorithm. Therefore, in this paper we use an interval logic based method,\nnamed the Duration Calculus (DC), to specify a simplified single-cell model of\nthe DCA. Based on the DC specifications with further induction, we find that\neach individual cell in the DCA can perform its function as a detector in\nreal-time. Since the DCA can be seen as many such cells operating in parallel,\nit is potentially capable of performing real-time detection. However, the\nanalysis process of the standard DCA constricts its real-time capability. As a\nresult, we conclude that the analysis process of the standard DCA should be\nreplaced by a real-time analysis component, which can perform periodic analysis\nfor the purpose of real-time detection.\n", " We construct bijections giving three \"codes\" for trees. These codes follow\nnaturally from the Matrix Tree Theorem of Tutte and have many advantages over\nthe one produced by Prufer in 1918. One algorithm gives explicitly a bijection\nthat is implicit in Orlin's manipulatorial proof of Cayley's formula (the\nformula was actually found first by Borchardt). Another is based on a proof of\nKnuth. The third is an implementation of Joyal's pseudo-bijective proof of the\nformula, and is equivalent to one previously found by Egecioglu and Remmel. In\neach case, we have at least two algorithms, one of which involves hands-on\nmanipulations of the tree while the other involves a combinatorial and linear\nalgebraic manipulation of a matrix.\n", " In this paper we obtain necessary and sufficient conditions for a linear\nbounded operator in a Hilbert space $H$ to have a three-diagonal complex\nsymmetric matrix with non-zero elements on the first sub-diagonal in an\northonormal basis in $H$. It is shown that a set of all such operators is a\nproper subset of a set of all complex symmetric operators with a simple\nspectrum. Similar necessary and sufficient conditions are obtained for a linear\nbounded operator in $H$ to have a three-diagonal complex skew-symmetric matrix\nwith non-zero elements on the first sub-diagonal in an orthonormal basis in\n$H$.\n", " We demonstrate existence and uniqueness of Picard--Vessiot extensions\nsatisfying prescribed properties, for systems of linear differential equations\nover a field satisfying the same properties, under some closure assumptions on\nthe field of constants. An example includes the case of a equations over a real\nfield, with a real-closed field of constants. The result is obtained through a\nmodel theoretic approach.\n", " Let $\\mathcal{F}$ be a family of $3$-uniform linear hypergraphs. The linear\nTur\\'an number of $\\mathcal F$ is the maximum possible number of edges in a\n$3$-uniform linear hypergraph on $n$ vertices which contains no member of\n$\\mathcal{F}$ as a subhypergraph.\n In this paper we show that the linear Tur\\'an number of the five cycle $C_5$\n(in the Berge sense) is $\\frac{1}{3 \\sqrt3}n^{3/2}$ asymptotically. We also\nshow that the linear Tur\\'an number of the four cycle $C_4$ and $\\{C_3, C_4\\}$\nare equal asmptotically, which is a strengthening of a theorem of Lazebnik and\nVerstra\\\"ete.\n We establish a connection between the linear Tur\\'an number of the linear\ncycle of length $2k+1$ and the extremal number of edges in a graph of girth\nmore than $2k-2$. Combining our result and a theorem of Collier-Cartaino,\nGraber and Jiang, we obtain that the linear Tur\\'an number of the linear cycle\nof length $2k+1$ is $\\Theta(n^{1+\\frac{1}{k}})$ for $k = 2, 3, 4, 6$.\n", " Both the classical time-ordering and the Magnus expansion are well-known in\nthe context of linear initial value problems. Motivated by the noncommutativity\nbetween time-ordering and time derivation, and related problems raised recently\nin statistical physics, we introduce a generalization of the Magnus expansion.\nWhereas the classical expansion computes the logarithm of the evolution\noperator of a linear differential equation, our generalization addresses the\nsame problem, including however directly a non-trivial initial condition. As a\nby-product we recover a variant of the time ordering operation, known as\nT*-ordering. Eventually, placing our results in the general context of\nRota-Baxter algebras permits us to present them in a more natural algebraic\nsetting. It encompasses, for example, the case where one considers linear\ndifference equations instead of linear differential equations.\n", " Using the concept of projective systems for linear codes and elementary\nlinear algebra, we show that projective $[n,k]_q$ codes form a connected\nsubgraph in the Grassmann graph consisting of $k$-dimensional subspaces of an\n$n$-dimensional vector space over the $q$-element field.\n", " The purpose of this paper is to build a new bridge between category theory\nand a generalized probability theory known as noncommutative probability or\nquantum probability, which was originated as a mathematical framework for\nquantum theory, in terms of states as linear functionals defined on category\nalgebras. We clarify that category algebras can be considered as generalized\nmatrix algebras and that states on categories as linear functionals defined on\ncategory algebras turn out to be generalized of probability measures on sets as\ndiscrete categories. Moreover, by establishing a generalization of famous GNS\n(Gelfand-Naimark-Segal) construction, we obtain representations of category\nalgebras of $^{\\dagger}$-categories on certain generalized Hilbert spaces which\nwe call semi-Hilbert modules over rigs.\n", " In 1977, Yu. V. Matiyasevich proposed a formula expressing the chromatic\npolynomial of an arbitrary graph as a linear combination of flow polynomials of\nsubgraphs of the original graph. In this paper, we prove that this\nrepresentation is a particular case of one (easily verifiable) formula, namely,\nthe representation of the characteristic polynomial of an arbitrary matroid as\na linear combination of characteristic polynomials of dual matroids. As an\napplication, we represent the flow polynomial of a complete graph with $n$\nvertices as the sum of elementary terms with respect to all partitions of\npositive integer $n$. Since the growth rate of the number of all partitions is\nless than exponential, this technique allows us to evaluate the flow polynomial\nfor values of $n\\approx 50$. We also get an explicit expression for the\ncharacteristic polynomial of the matroid dual to the matroid of the projective\ngeometry over a finite field. We prove, in particular, that major coefficients\nof all these polynomials coincide with the beginning of the row in the Pascal\ntriangle, whose number equals the quantity of elements in the corresponding\nmatroid. At the end part of the paper, we consider one more approach, which\nallows us to obtain the same results of application of our main theoren by\nusing properties of the Tutte polynomial and the classical Rota formula for\ncoefficients of the characteristic polynomial of a matroid. In addition, we\ndescribe the connection between the matroid variant of the Matiyasevich formula\nand convolution formulas for Tutte polynomials.\n", " Quantitatively predictive models of biomolecular circuits are important tools\nfor the design of synthetic biology and molecular communication circuits. The\ninformation content of typical time-lapse single-cell data for the inference of\nkinetic parameters is not only limited by measurement uncertainty and intrinsic\nstochasticity, but also by the employed perturbations. Novel microfluidic\ndevices enable the synthesis of temporal chemical concentration profiles. The\ninformativeness of a perturbation can be quantified based on mutual\ninformation. We propose an approximate method to perform optimal experimental\ndesign of such perturbation profiles. To estimate the mutual information we\nperform a multivariate log-normal approximation of the joint distribution over\nparameters and observations and scan the design space using Metropolis-Hastings\nsampling. The method is demonstrated by finding optimal perturbation sequences\nfor synthetic case studies on a gene expression model with varying reporter\ncharacteristics.\n", " Some Gruss type inequalities in semi-inner product modules over C*-algebras\nfor n-tuples of vectors are established. Also we give their natu- ral\napplications for the approximation of the discrete Fourier and the Melin\ntransforms of bounded linear operators on a Hilbert space.\n", " Gene transcription is a highly stochastic and dynamic process. As a result,\nthe mRNA copy number of a given gene is heterogeneous both between cells and\nacross time. We present a framework to model gene transcription in populations\nof cells with time-varying (stochastic or deterministic) transcription and\ndegradation rates. Such rates can be understood as upstream cellular drives\nrepresenting the effect of different aspects of the cellular environment. We\nshow that the full solution of the master equation contains two components: a\nmodel-specific, upstream effective drive, which encapsulates the effect of\ncellular drives (e.g., entrainment, periodicity or promoter randomness), and a\ndownstream transcriptional Poissonian part, which is common to all models. Our\nanalytical framework treats cell-to-cell and dynamic variability consistently,\nunifying several approaches in the literature. We apply the obtained solution\nto characterise different models of experimental relevance, and to explain the\ninfluence on gene transcription of synchrony, stationarity, ergodicity, as well\nas the effect of time-scales and other dynamic characteristics of drives. We\nalso show how the solution can be applied to the analysis of noise sources in\nsingle-cell data, and to reduce the computational cost of stochastic\nsimulations.\n", " We investigate the computational complexity of tensor rank, a concept that\nplays fundamental role in different topics of modern applied mathematics. For\ntensors over any integral domain, we prove that the rank problem is polynomial\ntime equivalent to solving a system of polynomial equations over this integral\ndomain. Our result gives a complete description of the algorithmic complexity\nof tensor rank and allows one to solve several known open problems. In\nparticular, the tensor rank over $\\mathbb{Z}$ turns out to be undecidable,\nwhich answers the question posed by Gonzalez and Ja'Ja' in 1980. We generalize\nour result and prove that the symmetric rank admits a similar description of\ncomputational complexity as the one we give for usual rank. In particular,\ncomputing the symmetric rank of a rational tensor is shown to be NP-hard, which\nproves a recent conjecture of Hillar and Lim. As a byproduct of our approach,\nwe get a similar characterization of the algorithmic complexity of the minimal\nrank matrix completion problem, which gives a complete answer to the question\ndiscussed in 1999 by Buss, Frandsen, and Shallit.\n", " The Escherichia coli chemoreceptors form an extensive array that achieves\ncooperative and adaptive sensing of extracellular signals. The receptors\ncontrol the activity of histidine kinase CheA, which drives a non-equilibrium\nphosphorylation-dephosphorylation reaction cycle for response regulator CheY.\nRecent single-cell FRET measurements revealed that kinase activity of the array\nspontaneously switches between active and inactive states, with asymmetric\nswitching times that signify time-reversal symmetry breaking in the underlying\ndynamics. Here, we show that the asymmetric switching dynamics can be explained\nby a non-equilibrium lattice model, which considers both the dissipative\nreaction cycles of individual core units and the coupling between neighboring\nunits. The model reveals that large dissipation and near-critical coupling are\nrequired to explain the observed switching dynamics. Microscopically, the\nswitching time asymmetry originates from irreversible transition paths. The\nmodel shows that strong dissipation enables sensitive and rapid signaling\nresponse by relieving the speed-sensitivity trade-off, which can be tested by\nfuture single-cell experiments. Overall, our model provides a general framework\nfor studying biological complexes composed of coupled subunits that are\nindividually driven by dissipative cycles and the rich non-equilibrium physics\nwithin.\n", " A linear equation Au=f (1) with a bounded, injective, but not boundedly\ninvertible linear operator in a Hilbert space H is studied. A new approach to\nsolving linear ill-posed problems is proposed. The approach consists of solving\na Cauchy problem for a linear equation in H, which is a dynamical system,\nproving the existence and uniqueness of its global solution u(t), and\nestablishing that u(t) tends to a limit y, as t tends to infinity, and this\nlimit y solves equation (1). The case when f in (1) is given with some error is\nalso studied.\n", " An embedding of a graph in a book, called book embedding, consists of a\nlinear ordering of its vertices along the spine of the book and an assignment\nof its edges to the pages of the book, so that no two edges on the same page\ncross. The book thickness of a graph is the minimum number of pages over all\nits book embeddings. For planar graphs, a fundamental result is due to\nYannakakis, who proposed an algorithm to compute embeddings of planar graphs in\nbooks with four pages. Our main contribution is a technique that generalizes\nthis result to a much wider family of nonplanar graphs, which is characterized\nby a biconnected skeleton of crossing-free edges whose faces have bounded\ndegree. Notably, this family includes all 1-planar, all optimal 2-planar, and\nall k-map (with bounded k) graphs as subgraphs. We prove that this family of\ngraphs has bounded book thickness, and as a corollary, we obtain the first\nconstant upper bound for the book thickness of optimal 2-planar and k-map\ngraphs.\n", " We propose a linear time and constant space algorithm for computing Euclidean\nprojections onto sets on which a normalized sparseness measure attains a\nconstant value. These non-convex target sets can be characterized as\nintersections of a simplex and a hypersphere. Some previous methods required\nthe vector to be projected to be sorted, resulting in at least quasilinear time\ncomplexity and linear space complexity. We improve on this by adaptation of a\nlinear time algorithm for projecting onto simplexes. In conclusion, we propose\nan efficient algorithm for computing the product of the gradient of the\nprojection with an arbitrary vector.\n", " Flows over time generalize classical network flows by introducing a notion of\ntime. Each arc is equipped with a transit time that specifies how long flow\ntakes to traverse it, while flow rates may vary over time within the given edge\ncapacities. In this paper, we extend this concept of a dynamic optimization\nproblem to the more general setting of abstract flows. In this model, the\nunderlying network is replaced by an abstract system of linearly ordered sets,\ncalled \"paths\" satisfying a simple switching property: Whenever two paths P and\nQ intersect, there must be another path that is contained in the beginning of P\nand the end of Q.\n We show that a maximum abstract flow over time can be obtained by solving a\nweighted abstract flow problem and constructing a temporally repeated flow from\nits solution. In the course of the proof, we also show that the relatively\nmodest switching property of abstract networks already captures many essential\nproperties of classical networks.\n", " Given a set of points $P$ and a set of regions $\\mathcal{O}$, an incidence is\na pair $(p,o ) \\in P \\times \\mathcal{O}$ such that $p \\in o$. We obtain a\nnumber of new results on a classical question in combinatorial geometry: What\nis the number of incidences (under certain restrictive conditions)?\n We prove a bound of $O\\bigl( k n(\\log n/\\log\\log n)^{d-1} \\bigr)$ on the\nnumber of incidences between $n$ points and $n$ axis-parallel boxes in\n$\\mathbb{R}^d$, if no $k$ boxes contain $k$ common points, that is, if the\nincidence graph between the points and the boxes does not contain $K_{k,k}$ as\na subgraph. This new bound improves over previous work, by Basit, Chernikov,\nStarchenko, Tao, and Tran (2021), by more than a factor of $\\log^d n$ for $d\n>2$. Furthermore, it matches a lower bound implied by the work of Chazelle\n(1990), for $k=2$, thus settling the question for points and boxes.\n We also study several other variants of the problem. For halfspaces, using\nshallow cuttings, we get a linear bound in two and three dimensions. We also\npresent linear (or near linear) bounds for shapes with low union complexity,\nsuch as pseudodisks and fat triangles.\n", " For any site of definition $\\mathcal C$ of a Grothendieck topos $\\mathcal E$,\nwe define a notion of a $\\mathcal C$-ary Lawvere theory $\\tau: \\mathscr C \\to\n\\mathscr T$ whose category of models is a stack over $\\mathcal E$. Our\ndefinitions coincide with Lawvere's finitary theories when $\\mathcal\nC=\\aleph_0$ and $\\mathcal E = \\operatorname{\\mathbf {Set}}$. We construct a\nfibered category $\\operatorname{\\mathbf {Mod}}^{\\mathscr T}$ of models as a\nstack over $\\mathcal E$ and prove that it is $\\mathcal E$-complete and\n$\\mathcal E$-cocomplete. We show that there is a free-forget adjunction $F\n\\dashv U: \\operatorname{\\mathbf {Mod}}^{\\mathscr T} \\rightleftarrows \\mathscr\nE$. If $\\tau$ is a commutative theory in a certain sense, then we obtain a\n``locally monoidal closed'' structure on the category of models, which enhances\nthe free-forget adjunction to an adjunction of symmetric monoidal $\\mathcal\nE$-categories. Our results give a general recipe for constructing a monoidal\n$\\mathcal E$-cosmos in which one can do enriched $\\mathcal E$-category theory.\nAs an application, we describe a convenient category of linear spaces generated\nby the theory of Lebesgue integration.\n", " Systems of polynomial equations over an algebraically-closed field K can be\nused to concisely model many combinatorial problems. In this way, a\ncombinatorial problem is feasible (e.g., a graph is 3-colorable, hamiltonian,\netc.) if and only if a related system of polynomial equations has a solution\nover K. In this paper, we investigate an algorithm aimed at proving\ncombinatorial infeasibility based on the observed low degree of Hilbert's\nNullstellensatz certificates for polynomial systems arising in combinatorics\nand on large-scale linear-algebra computations over K. We report on experiments\nbased on the problem of proving the non-3-colorability of graphs. We\nsuccessfully solved graph problem instances having thousands of nodes and tens\nof thousands of edges.\n", " A planar point set is in convex position precisely when it has a convex\npolygonization, that is, a polygonization with maximum interior angle measure\nat most \\pi. We can thus talk about the convexity of a set of points in terms\nof the minimum, taken over all polygonizations, of the maximum interior angle.\nThe main result presented here is a nontrivial combinatorial upper bound of\nthis min-max value in terms of the number of points in the set. Motivated by a\nparticular construction, we also pose a natural conjecture for the best upper\nbound.\n", " We briefly review the problems and prospects of the standard lore of dark\nenergy and notice that the cosmological constant problems are often over\nemphasized. In order to keep the discussion pedagogical aimed at a wider\naudience, we have avoided technical complications in several places and\nresorted to heuristic arguments based on physical perceptions. We presented\nunderlying ideas of modified theories based upon chameleon mechanism and\nVainshtein screening. We have given a lucid illustration of recently\ninvestigated ghost free non linear massive gravity. Again we have sacrificed\nrigor and confined to the basic ideas that led to the formulation of $dRGT$.\nThe review ends with a brief discussion on the difficulties of the theory\napplied to cosmology with an optimism that a consistent solution of the problem\nis round the corner.\n", " Let M be a matroid representable over a (partial) field P and B a matrix\nrepresentable over a sub-partial field P' of P. We say that B confines M to P'\nif, whenever a P-representation matrix A of M has a submatrix B, A is a scaled\nP'-matrix. We show that, under some conditions on the partial fields, on M, and\non B, verifying whether B confines M to P' amounts to a finite check. A\ncorollary of this result is Whittle's Stabilizer Theorem.\n A combination of the Confinement Theorem and the Lift Theorem from\narXiv:0804.3263 leads to a short proof of Whittle's characterization of the\nmatroids representable over GF(3) and other fields.\n We also use a combination of the Confinement Theorem and the Lift Theorem to\nprove a characterization, in terms of representability over partial fields, of\nthe 3-connected matroids that have k inequivalent representations over GF(5),\nfor k = 1, ..., 6.\n Additionally we give, for a fixed matroid M, an algebraic construction of a\npartial field P_M and a representation A over P_M such that every\nrepresentation of M over a partial field P is equal to f(A) for some\nhomomorphism f:P_M->P. Using the Confinement Theorem we prove an algebraic\nanalog of the theory of free expansions by Geelen et al.\n", " A (multi)set of segments in the plane may form a TSP tour, a matching, a\ntree, or any multigraph. If two segments cross, then we can reduce the total\nlength with the following flip operation. We remove a pair of crossing\nsegments, and insert a pair of non-crossing segments, while keeping the same\nvertex degrees. The goal of this paper is to devise efficient strategies to\nflip the segments in order to obtain crossing-free segments after a small\nnumber of flips. Linear and near-linear bounds on the number of flips were only\nknown for segments with endpoints in convex position. We generalize these\nresults, proving linear and near-linear bounds for cases with endpoints that\nare not in convex position. Our results are proved in a general setting that\napplies to multiple problems, using multigraphs and the distinction between\nremoval and insertion choices when performing a flip.\n", " Let $A$ be a right linear operator on a two-sided quaternionic Banach space\n$X$. The paper studies the Drazin inverse for right linear operators on a\nquaternionic Banach space. It is shown that if $A$ is Drazin invertible then\nthe Drazin inverse of $A$ is given by $f(A)$ where $f$ is $0$ in an axially\nsymmetric neighborhood of $0$ and $f(q) = q^{-1}$ in an axially symmetric\nneighborhood of the nonzero spherical spectrum of $A$. Some results analogous\nto the ones concerning the Drazin inverse of operators on complex Banach spaces\nare proved in the quaternionic context.\n", " The minrank of a graph $G$ on the set of vertices $[n]$ over a field\n$\\mathbb{F}$ is the minimum possible rank of a matrix $M\\in\\mathbb{F}^{n\\times\nn}$ with nonzero diagonal entries such that $M_{i,j}=0$ whenever $i$ and $j$\nare distinct nonadjacent vertices of $G$. This notion, over the real field,\narises in the study of the Lov\\'asz theta function of a graph. We obtain tight\nbounds for the typical minrank of the binomial random graph $G(n,p)$ over any\nfinite or infinite field, showing that for every field $\\mathbb{F}=\\mathbb\nF(n)$ and every $p=p(n)$ satisfying $n^{-1} \\leq p \\leq 1-n^{-0.99}$, the\nminrank of $G=G(n,p)$ over $\\mathbb{F}$ is $\\Theta(\\frac{n \\log (1/p)}{\\log\nn})$ with high probability. The result for the real field settles a problem\nraised by Knuth in 1994. The proof combines a recent argument of Golovnev,\nRegev, and Weinstein, who proved the above result for finite fields of size at\nmost $n^{O(1)}$, with tools from linear algebra, including an estimate of\nR\\'onyai, Babai, and Ganapathy for the number of zero-patterns of a sequence of\npolynomials.\n", " The problem involving the extension of functions from a certain class and\ndefined on subdomains of the ambient space to the whole space is an old and a\nwell investigated theme in analysis. A related question whether the extensions\nthat result in the process may be chosen in a linear or a continuous manner\nbetween appropriate spaces of functions turns out to be highly nontrivial. That\nthis holds for the class of continuous functions defined on metric spaces is\nthe well-known Borsuk-Dugundji theorem which asserts that given a metric space\nM and a subspace S of M, each continuous function g on S can be extended to a\ncontinuous function f on X such that the resulting assignment from C(S) to C(M)\nis a norm-one continuous linear extension operator.\n The present paper is devoted to an investigation of this problem in the\ncontext of extendability of Lipschitz functions from closed subspaces of a\ngiven Banach space to the whole space such that the choice of the extended\nfunction gives rise to a bounded linear (extension) operator between\nappropriate spaces of Lipschitz functions. It is shown that the indicated\nproperty holds precisely when the underlying space is isomorphic to a Hilbert\nspace. Among certain useful consequences of this theorem, we provide an\nisomorphic analogue of a well-known theorem of S. Reich by show ing that closed\nconvex subsets of a Banach space X arise as Lipschitz retracts of X precisely\nwhen X is isomorphically a Hilbert space. We shall also discuss the issue of\nbounded linear extension operators between spaces of Lipschitz functions now\ndefined on arbitrary subsets of Banach spaces and provide a direct proof of the\nknown non-existence of such an extension operator by using methods which are\nmore accessible than those initially employed by the authors.\n", " In this work, we study discontinuous Sturm-Liouville type problems with\neigenparameter dependent boundary condition and transmission conditions at\nthree interior points. A self-adjoint linear operator A is defined in a\nsuitable Hilbert space H such that the eigenvalues of such a problem coincide\nwith those of A. We show that the eigenvalues of the problem are analytically\nsimple, and the eigenfunctions of A are complete in H.\n", " Given a finite set $X \\subset \\mathbb{R}^d$ and a binary linear classifier\n$c: \\mathbb{R}^d \\to \\{0,1\\}$, how many queries of the form $c(x)$ are required\nto learn the label of every point in $X$? Known as \\textit{point location},\nthis problem has inspired over 35 years of research in the pursuit of an\noptimal algorithm. Building on the prior work of Kane, Lovett, and Moran (ICALP\n2018), we provide the first nearly optimal solution, a randomized linear\ndecision tree of depth $\\tilde{O}(d\\log(|X|))$, improving on the previous best\nof $\\tilde{O}(d^2\\log(|X|))$ from Ezra and Sharir (Discrete and Computational\nGeometry, 2019). As a corollary, we also provide the first nearly optimal\nalgorithm for actively learning halfspaces in the membership query model. En\nroute to these results, we prove a novel characterization of Barthe's Theorem\n(Inventiones Mathematicae, 1998) of independent interest. In particular, we\nshow that $X$ may be transformed into approximate isotropic position if and\nonly if there exists no $k$-dimensional subspace with more than a\n$k/d$-fraction of $X$, and provide a similar characterization for exact\nisotropic position.\n", " The computation of determinants or their signs is the core procedure in many\nimportant geometric algorithms, such as convex hull, volume and point location.\nAs the dimension of the computation space grows, a higher percentage of the\ntotal computation time is consumed by these computations. In this paper we\nstudy the sequences of determinants that appear in geometric algorithms. The\ncomputation of a single determinant is accelerated by using the information\nfrom the previous computations in that sequence.\n We propose two dynamic determinant algorithms with quadratic arithmetic\ncomplexity when employed in convex hull and volume computations, and with\nlinear arithmetic complexity when used in point location problems. We implement\nthe proposed algorithms and perform an extensive experimental analysis. On one\nhand, our analysis serves as a performance study of state-of-the-art\ndeterminant algorithms and implementations. On the other hand, we demonstrate\nthe supremacy of our methods over state-of-the-art implementations of\ndeterminant and geometric algorithms. Our experimental results include a 20 and\n78 times speed-up in volume and point location computations in dimension 6 and\n11 respectively.\n", " In this paper we show that under suitable conditions on their Fourier--Bohr\ncoefficients, the twisted Eberlein convolution of a measure with pure point\ndiffraction spectra and a measure with continuous diffraction spectra is zero.\nIn particular, the diffraction spectrum of a linear combinations of the two\nmeasures is simply the linear combinations of the two diffraction spectra with\nabsolute value square coefficients.\n", " We provide a new, short proof of the density in energy of Lipschitz functions\ninto the metric Sobolev space defined by using plans with barycenter (and thus,\na fortiori, into the Newtonian-Sobolev space). Our result covers first-order\nSobolev spaces of exponent $p\\in(1,\\infty)$, defined over a complete and\nseparable metric space endowed with a boundedly-finite Borel measure.\n Our proof is based on a completely smooth analysis: first we reduce the\nproblem to the Banach space setting, where we consider smooth functions instead\nof Lipschitz ones, then we rely on classical tools in convex analysis and on\nthe superposition principle for normal $1$-currents. Along the way, we obtain a\nnew proof of the density in energy of smooth cylindrical functions in Sobolev\nspaces defined over a separable Banach space endowed with a finite Borel\nmeasure.\n", " We give examples showing that the usual Artin Approximation theorems valid\nfor convergent series over a field are no longer true for convergent series\nover a commutative Banach algebra. In particular we construct an example of a\ncommutative integral Banach algebra $R$ such that the ring of formal power\nseries over $R$ is not flat over the ring of convergent power series over $R$.\n", " Polytope numbers for a given polytope are an integer sequence defined by the\ncombinatorics of the polytope. Recent work by H. K. Kim and J. Y. Lee has\nfocused on writing polytope number sequences as sums of simplex number\nsequences. In addition, these works have given a process for writing the\npolytope number sequence in a recursive fashion by using the interior sequence\nfor the various k-faces of the polytope, each viewed as a k-dimensional\npolytope. This paper shows that the coefficients of the linear combination of\nsimplex number are the h-vector components for a certain type of triangulation\nof the polytope. In addition, reversing the order of the coefficients in the\nlinear combination is shown to equal the interior polytope sequence for this\npolytope.\n", " We show that the multivariate generating function of appropriately normalized\nmoments of a measure with homogeneous polynomial density supported on a compact\npolytope P in R^d is a rational function. Its denominator is the product of\nlinear forms dual to the vertices of P raised to the power equal to the degree\nof the density function. Using this, we solve the inverse moment problem for\nthe set of, not necessarily convex, polytopes having a given set S of vertices.\nUnder a weak non-degeneracy assumption we also show that the uniform measure\nsupported on any such polytope is a linear combination of uniform measures\nsupported on simplices with vertices in S.\n", " A classic theorem by Steinitz states that a graph G is realizable by a convex\npolyhedron if and only if G is 3-connected planar. Zonohedra are an important\nsubclass of convex polyhedra having the property that the faces of a zonohedron\nare parallelograms and are in parallel pairs. In this paper we give\ncharacterization of graphs of zonohedra. We also give a linear time algorithm\nto recognize such a graph. In our quest for finding the algorithm, we prove\nthat in a zonohedron P both the number of zones and the number of faces in each\nzone is O(square root{n}), where n is the number of vertices of P.\n", " We prove a differential analogue of Hilbert's irreducibility theorem. Let\n$\\mathcal{L}$ be a linear differential operator with coefficients in\n$C(\\mathbb{X})(x)$ that is irreducible over $\\overline{C(\\mathbb{X})}(x)$,\nwhere $\\mathbb{X}$ is an irreducible affine algebraic variety over an\nalgebraically closed field $C$ of characteristic zero. We show that the set of\n$c\\in \\mathbb{X}(C)$ such that the specialized operator $\\mathcal{L}^c$ of\n$\\mathcal{L}$ remains irreducible over $C(x)$ is Zariski dense in\n$\\mathbb{X}(C)$.\n" ]

[ " We propose a method that morphs high-orger meshes such that their boundaries\nand interfaces coincide/align with implicitly defined geometries. Our focus is\nparticularly on the case when the target surface is prescribed as the zero\nisocontour of a smooth discrete function. Common examples of this scenario\ninclude using level set functions to represent material interfaces in\nmultimaterial configurations, and evolving geometries in shape and topology\noptimization. The proposed method formulates the mesh optimization problem as a\nvariational minimization of the sum of a chosen mesh-quality metric using the\nTarget-Matrix Optimization Paradigm (TMOP) and a penalty term that weakly\nforces the selected faces of the mesh to align with the target surface. The\ndistinct features of the method are use of a source mesh to represent the level\nset function with sufficient accuracy, and adaptive strategies for setting the\npenalization weight and selecting the faces of the mesh to be fit to the target\nisocontour of the level set field. We demonstrate that the proposed method is\nrobust for generating boundary- and interface-fitted meshes for curvilinear\ndomains using different element types in 2D and 3D.\n", " We investigate exact crossing minimization for graphs that differ from trees\nby a small number of additional edges, for several variants of the crossing\nminimization problem. In particular, we provide fixed parameter tractable\nalgorithms for the 1-page book crossing number, the 2-page book crossing\nnumber, and the minimum number of crossed edges in 1-page and 2-page book\ndrawings.\n", " We prove the first nontrivial worst-case lower bounds for two closely related\nproblems. First, $\\Omega(n^{3/2})$ degree-1 reductions, series-parallel\nreductions, and $\\Delta$Y transformations are required in the worst case to\nreduce an $n$-vertex plane graph to a single vertex or edge. The lower bound is\nachieved by any planar graph with treewidth $\\Theta(\\sqrt{n})$. Second,\n$\\Omega(n^{3/2})$ homotopy moves are required in the worst case to reduce a\nclosed curve in the plane with $n$ self-intersection points to a simple closed\ncurve. For both problems, the best upper bound known is $O(n^2)$, and the only\nlower bound previously known was the trivial $\\Omega(n)$.\n The first lower bound follows from the second using medial graph techniques\nultimately due to Steinitz, together with more recent arguments of Noble and\nWelsh [J. Graph Theory 2000]. The lower bound on homotopy moves follows from an\nobservation by Haiyashi et al. [J. Knot Theory Ramif. 2012] that the standard\nprojections of certain torus knots have large defect, a topological invariant\nof generic closed curves introduced by Aicardi and Arnold. Finally, we prove\nthat every closed curve in the plane with $n$ crossings has defect\n$O(n^{3/2})$, which implies that better lower bounds for our algorithmic\nproblems will require different techniques.\n", " Estimating the number of vertices of a two dimensional projection, called a\nshadow, of a polytope is a fundamental tool for understanding the performance\nof the shadow simplex method for linear programming among other applications.\nWe prove multiple upper bounds on the expected number of vertices of a random\nshadow of a fixed polytope. Our bounds are in terms of various parameters in\nthe literature including geometric diameter and edge lengths, minimal and\nmaximal slack, maximal coordinates for lattice polytopes, and maximum absolute\nvalues of subdeterminants. For the case of geometric diameter and edge lengths,\nwe prove lower bounds and argue that our upper and lower bounds are both tight\nfor zonotopes.\n", " Many biological processes discriminate between correct and incorrect\nsubstrates through the kinetic proofreading mechanism which enables lower error\nat the cost of higher energy dissipation. Elucidating physicochemical\nconstraints for global minimization of dissipation and error is important for\nunderstanding enzyme evolution. Here, we identify theoretically a fundamental\nerror-cost bound which tightly constrains the performance of proofreading\nnetworks under any parameter variations preserving the rate discrimination\nbetween substrates. The bound is kinetically controlled, i.e. completely\ndetermined by the difference between the transition state energies on the\nunderlying free energy landscape. The importance of the bound is analyzed for\nthree biological processes. DNA replication by T7 DNA polymerase is shown to be\nnearly optimized, i.e. its kinetic parameters place it in the immediate\nproximity of the error-cost bound. The isoleucyl-tRNA synthetase (IleRS) of E.\ncoli also operates close to the bound, but further optimization is prevented by\nthe need for reaction speed. In contrast, E. coli ribosome operates in a\nhigh-dissipation regime, potentially in order to speed up protein production.\nTogether, these findings establish a fundamental error-dissipation relation in\nbiological proofreading networks and provide a theoretical framework for\nstudying error-dissipation trade-off in other systems with biological\ndiscrimination.\n", " We give improved pseudorandom generators (PRGs) for Lipschitz functions of\nlow-degree polynomials over the hypercube. These are functions of the form\npsi(P(x)), where P is a low-degree polynomial and psi is a function with small\nLipschitz constant. PRGs for smooth functions of low-degree polynomials have\nreceived a lot of attention recently and play an important role in constructing\nPRGs for the natural class of polynomial threshold functions. In spite of the\nrecent progress, no nontrivial PRGs were known for fooling Lipschitz functions\nof degree O(log n) polynomials even for constant error rate. In this work, we\ngive the first such generator obtaining a seed-length of (log\nn)\\tilde{O}(d^2/eps^2) for fooling degree d polynomials with error eps.\nPrevious generators had an exponential dependence on the degree.\n We use our PRG to get better integrality gap instances for sparsest cut, a\nfundamental problem in graph theory with many applications in graph\noptimization. We give an instance of uniform sparsest cut for which a powerful\nsemi-definite relaxation (SDP) first introduced by Goemans and Linial and\nstudied in the seminal work of Arora, Rao and Vazirani has an integrality gap\nof exp(\\Omega((log log n)^{1/2})). Understanding the performance of the\nGoemans-Linial SDP for uniform sparsest cut is an important open problem in\napproximation algorithms and metric embeddings and our work gives a\nnear-exponential improvement over previous lower bounds which achieved a gap of\n\\Omega(log log n).\n", " Weak and strong coloring numbers are generalizations of the degeneracy of a\ngraph, where for each natural number $k$, we seek a vertex ordering such every\nvertex can (weakly respectively strongly) reach in $k$ steps only few vertices\nwith lower index in the ordering. Both notions capture the sparsity of a graph\nor a graph class, and have interesting applications in the structural and\nalgorithmic graph theory. Recently, the first author together with McCarty and\nNorin observed a natural volume-based upper bound for the strong coloring\nnumbers of intersection graphs of well-behaved objects in $\\mathbb{R}^d$, such\nas homothets of a centrally symmetric compact convex object, or comparable\naxis-aligned boxes.\n In this paper, we prove upper and lower bounds for the $k$-th weak coloring\nnumbers of these classes of intersection graphs. As a consequence, we describe\na natural graph class whose strong coloring numbers are polynomial in $k$, but\nthe weak coloring numbers are exponential. We also observe a surprising\ndifference in terms of the dependence of the weak coloring numbers on the\ndimension between touching graphs of balls (single-exponential) and hypercubes\n(double-exponential).\n", " We show conditional lower bounds for well-studied #P-hard problems:\n (a) The number of satisfying assignments of a 2-CNF formula with n variables\ncannot be counted in time exp(o(n)), and the same is true for computing the\nnumber of all independent sets in an n-vertex graph.\n (b) The permanent of an n x n matrix with entries 0 and 1 cannot be computed\nin time exp(o(n)).\n (c) The Tutte polynomial of an n-vertex multigraph cannot be computed in time\nexp(o(n)) at most evaluation points (x,y) in the case of multigraphs, and it\ncannot be computed in time exp(o(n/polylog n)) in the case of simple graphs.\n Our lower bounds are relative to (variants of) the Exponential Time\nHypothesis (ETH), which says that the satisfiability of n-variable 3-CNF\nformulas cannot be decided in time exp(o(n)). We relax this hypothesis by\nintroducing its counting version #ETH, namely that the satisfying assignments\ncannot be counted in time exp(o(n)). In order to use #ETH for our lower bounds,\nwe transfer the sparsification lemma for d-CNF formulas to the counting\nsetting.\n", " We use the framework of the first-order differential structure in metric\nmeasure spaces introduced by Gigli to define a notion of weak solutions to\ngradient flows of convex, lower semicontinuous and coercive functionals. We\nprove their existence and uniqueness and show that they are also variational\nsolutions; in particular, this is an existence result for variational\nsolutions. Then, we apply this technique in the case of a gradient flow of a\nfunctional with inhomogeneous growth.\n", " Tusn\\'ady's problem asks to bound the discrepancy of points and axis-parallel\nboxes in $\\mathbb{R}^d$. Algorithmic bounds on Tusn\\'ady's problem use a\ncanonical decomposition of Matou\\v{s}ek for the system of points and\naxis-parallel boxes, together with other techniques like partial coloring and /\nor random-walk based methods. We use the notion of \\emph{shallow cell\ncomplexity} and the \\emph{shallow packing lemma}, together with the chaining\ntechnique, to obtain an improved decomposition of the set system. Coupled with\nan algorithmic technique of Bansal and Garg for discrepancy minimization, which\nwe also slightly extend, this yields improved algorithmic bounds on Tusn\\'ady's\nproblem. For $d\\geq 5$, our bound matches the lower bound of\n$\\Omega(\\log^{d-1}n)$ given by Matou\\v{s}ek, Nikolov and Talwar [IMRN, 2020] --\nsettling Tusn\\'ady's problem, upto constant factors. For $d=2,3,4$, we obtain\nimproved algorithmic bounds of $O(\\log^{7/4}n)$, $O(\\log^{5/2}n)$ and\n$O(\\log^{13/4}n)$ respectively, which match or improve upon the\nnon-constructive bounds of Nikolov for $d\\geq 3$. Further, we also give\nimproved bounds for the discrepancy of set systems of points and polytopes in\n$\\mathbb{R}^d$ generated via translations of a fixed set of hyperplanes. As an\napplication, we also get a bound for the geometric discrepancy of anchored\nboxes in $\\mathbb{R}^d$ with respect to an arbitrary measure, matching the\nupper bound for the Lebesgue measure, which improves on a result of\nAistleitner, Bilyk, and Nikolov [MC and QMC methods, \\emph{Springer, Proc.\nMath. Stat.}, 2018] for $d\\geq 4$.\n", " In three-dimensional computational topology, the theory of normal surfaces is\na tool of great theoretical and practical significance. Although this theory\ntypically leads to exponential time algorithms, very little is known about how\nthese algorithms perform in \"typical\" scenarios, or how far the best known\ntheoretical bounds are from the real worst-case scenarios. Here we study the\ncombinatorial and algebraic complexity of normal surfaces from both the\ntheoretical and experimental viewpoints. Theoretically, we obtain new\nexponential lower bounds on the worst-case complexities in a variety of\nsettings that are important for practical computation. Experimentally, we study\nthe worst-case and average-case complexities over a comprehensive body of\nroughly three billion input triangulations. Many of our lower bounds are the\nfirst known exponential lower bounds in these settings, and experimental\nevidence suggests that many of our theoretical lower bounds on worst-case\ngrowth rates may indeed be asymptotically tight.\n", " Given two point sets in the plane, we study the minimization of the\nbottleneck distance between a point set B and an equally-sized subset of a\npoint set A under translations. We relate this problem to a Voronoi-type\ndiagram and derive polynomial bounds for its complexity that are optimal in the\nsize of A. We devise efficient algorithms for the construction of such a\ndiagram and its lexicographic variant, which generalize to higher dimensions.\nWe use the diagram to find an optimal bottleneck matching under translations,\nto compute a connecting path of minimum bottleneck cost between two positions\nof B, and to determine the maximum bottleneck cost in a convex polygon.\n", " A space $X$ is $n$-arc connected (respectively, $n$-circle connected) if for\nany choice of at most $n$ points there is an arc (respectively, a circle) in\n$X$ containing the specified points. We study $n$-arc connectedness and\n$n$-circle connectedness in compactifications of locally finite graphs and the\nslightly more general class of graph-like continua, uncovering a striking\ndifference in their behaviour regarding $n$-arc and -circle connectedness.\n", " A metric graph is a 1-dimensional stratified metric space consisting of\nvertices and edges or loops glued together. Metric graphs can be naturally used\nto represent and model data that take the form of noisy filamentary structures,\nsuch as street maps, neurons, networks of rivers and galaxies. We consider the\nstatistical problem of reconstructing the topology of a metric graph embedded\nin R^D from a random sample. We derive lower and upper bounds on the minimax\nrisk for the noiseless case and tubular noise case. The upper bound is based on\nthe reconstruction algorithm given in Aanjaneya et al. (2012).\n", " A Newton graph of order $r( \\geqslant 2)$ is a cellularly embedded toroidal\ngraph on $r$ vertices, $2r$ edges and $r$ faces that fulfils certain\ncombinatorial properties (Euler, Hall). The significance of these graphs relies\non their role in the study of structurally stable elliptic Newton flows - say\n$\\bar{\\bar{\\mathcal{N}}} (f)$ - of order $r$, i.e. desingularized continuous\nversions of Newton's iteration method for finding zeros for an elliptic\nfunction $f$ (of order $r$). In previous work we established a representation\nof these flows in terms of Newton graphs. The present paper results into the\nclassification of all $3^{\\text{rd}}$ order Newton graphs, implying a list of\nall nine possible $3^{\\text{rd}}$ order flows $\\bar{\\bar{\\mathcal{N}}} (f)$ (up\nto conjugacy and duality).\n", " We study a class of optimization problems including matrix scaling, matrix\nbalancing, multidimensional array scaling, operator scaling, and tensor scaling\nthat arise frequently in theory and in practice. Some of these problems, such\nas matrix and array scaling, are convex in the Euclidean sense, but others such\nas operator scaling and tensor scaling are geodesically convex on a different\nRiemannian manifold. Trust region methods, which include box-constrained\nNewton's method, are known to produce high precision solutions very quickly for\nmatrix scaling and matrix balancing (Cohen et. al., FOCS 2017, Allen-Zhu et.\nal. FOCS 2017), and result in polynomial time algorithms for some geodesically\nconvex problems like operator scaling (Garg et. al. STOC 2018, B\\\"urgisser et.\nal. FOCS 2019). One is led to ask whether these guarantees also hold for\nmultidimensional array scaling and tensor scaling.\n We show that this is not the case by exhibiting instances with exponential\ndiameter bound: we construct polynomial-size instances of 3-dimensional array\nscaling and 3-tensor scaling whose approximate solutions all have doubly\nexponential condition number. Moreover, we study convex-geometric notions of\ncomplexity known as margin and gap, which are used to bound the running times\nof all existing optimization algorithms for such problems. We show that margin\nand gap are exponentially small for several problems including array scaling,\ntensor scaling and polynomial scaling. Our results suggest that it is\nimpossible to prove polynomial running time bounds for tensor scaling based on\ndiameter bounds alone. Therefore, our work motivates the search for analogues\nof more sophisticated algorithms, such as interior point methods, for\ngeodesically convex optimization that do not rely on polynomial diameter\nbounds.\n", " We investigate the problem of drawing graphs in 2D and 3D such that their\nedges (or only their vertices) can be covered by few lines or planes. We insist\non straight-line edges and crossing-free drawings. This problem has many\nconnections to other challenging graph-drawing problems such as small-area or\nsmall-volume drawings, layered or track drawings, and drawing graphs with low\nvisual complexity. While some facts about our problem are implicit in previous\nwork, this is the first treatment of the problem in its full generality. Our\ncontribution is as follows.\n We show lower and upper bounds for the numbers of lines and planes needed for\ncovering drawings of graphs in certain graph classes. In some cases our bounds\nare asymptotically tight; in some cases we are able to determine exact values.\n We relate our parameters to standard combinatorial characteristics of graphs\n(such as the chromatic number, treewidth, maximum degree, or arboricity) and to\nparameters that have been studied in graph drawing (such as the track number or\nthe number of segments appearing in a drawing).\n We pay special attention to planar graphs. For example, we show that there\nare planar graphs that can be drawn in 3-space on a lot fewer lines than in the\nplane.\n", " Prediction+optimization is a common real-world paradigm where we have to\npredict problem parameters before solving the optimization problem. However,\nthe criteria by which the prediction model is trained are often inconsistent\nwith the goal of the downstream optimization problem. Recently,\ndecision-focused prediction approaches, such as SPO+ and direct optimization,\nhave been proposed to fill this gap. However, they cannot directly handle the\nsoft constraints with the $max$ operator required in many real-world\nobjectives. This paper proposes a novel analytically differentiable surrogate\nobjective framework for real-world linear and semi-definite negative quadratic\nprogramming problems with soft linear and non-negative hard constraints. This\nframework gives the theoretical bounds on constraints' multipliers, and derives\nthe closed-form solution with respect to predictive parameters and thus\ngradients for any variable in the problem. We evaluate our method in three\napplications extended with soft constraints: synthetic linear programming,\nportfolio optimization, and resource provisioning, demonstrating that our\nmethod outperforms traditional two-staged methods and other decision-focused\napproaches.\n", " Let $G$ be a graph of a network system with vertices, $V(G)$, representing\nphysical locations and edges, $E(G)$, representing informational connectivity.\nA \\emph{locating-dominating (LD)} set $S$ is a subset of vertices representing\ndetectors capable of sensing an \"intruder\" at precisely their location or\nsomewhere in their open-neighborhood -- an LD set must be capable of locating\nan intruder anywhere in the graph. We explore three types of fault-tolerant LD\nsets: \\emph{redundant LD} sets, which allow a detector to be removed,\n\\emph{error-detecting LD} sets, which allow at most one false negative, and\n\\emph{error-correcting LD} sets, which allow at most one error (false positive\nor negative). In particular, we determine lower and upper bounds for the\nminimum density of fault-tolerant locating-dominating sets in the\n\\emph{infinite king grid}; to prove the lower bounds, we introduce a new\nshare-discharging strategy.\n", " Given a connected graph with domination (or total domination) number\n\\gamma>=2, we ask for the maximum number m_\\gamma and m_{\\gamma,T} of\ndominating and total dominating sets of size \\gamma. An exact answer is\nprovided for \\gamma=2and lower bounds are given for \\gamma>=3.\n", " Directed acyclic graph (DAG) models are popular for capturing causal\nrelationships. From observational and interventional data, a DAG model can only\nbe determined up to its \\emph{interventional Markov equivalence class} (I-MEC).\nWe investigate the size of MECs for random DAG models generated by uniformly\nsampling and ordering an Erd\\H{o}s-R\\'{e}nyi graph. For constant density, we\nshow that the expected $\\log$ observational MEC size asymptotically (in the\nnumber of vertices) approaches a constant. We characterize I-MEC size in a\nsimilar fashion in the above settings with high precision. We show that the\nasymptotic expected number of interventions required to fully identify a DAG is\na constant. These results are obtained by exploiting Meek rules and coupling\narguments to provide sharp upper and lower bounds on the asymptotic quantities,\nwhich are then calculated numerically up to high precision. Our results have\nimportant consequences for experimental design of interventions and the\ndevelopment of algorithms for causal inference.\n", " The availability of accurate localization is critical for multi-robot\nexploration strategies; noisy or inconsistent localization causes failure in\nmeeting exploration objectives. We aim to achieve high localization accuracy\nwith contemporary exploration map belief and vice versa without needing global\nlocalization information. This paper proposes a novel simultaneous exploration\nand localization (SEAL) approach, which uses Gaussian Processes (GP)-based\ninformation fusion for maximum exploration while performing communication graph\noptimization for relative localization. Both these cross-dependent objectives\nwere integrated through the Rao-Blackwellization technique. Distributed\nlinearized convex hull optimization is used to select the next-best unexplored\nregion for distributed exploration. SEAL outperformed cutting-edge methods on\nexploration and localization performance in extensive ROS-Gazebo simulations,\nillustrating the practicality of the approach in real-world applications.\n", " We present Higgs boson mass bounds in a lattice regularization allowing thus\nfor non-perturbative investigations. In particular, we employ a lattice\nmodified chiral invariant Higgs-Yukawa model using the overlap operator. We\nshow results for the upper and lower Higgs boson mass bounds in the presence of\na heavy mass-degenerate quark doublet with masses ranging up to 700 GeV. We\nperform infinite volume extrapolations in most cases, and examine several\nvalues of the lattice cutoff. Furthermore, we argue that the lower Higgs boson\nmass bound is stable with respect to the addition of higher dimensional\noperators to the scalar field potential. Our results have severe consequences\nfor the phenomenology of a fourth generation of quarks if a light Higgs boson\nis discovered at the LHC.\n", " Flows over time generalize classical network flows by introducing a notion of\ntime. Each arc is equipped with a transit time that specifies how long flow\ntakes to traverse it, while flow rates may vary over time within the given edge\ncapacities. In this paper, we extend this concept of a dynamic optimization\nproblem to the more general setting of abstract flows. In this model, the\nunderlying network is replaced by an abstract system of linearly ordered sets,\ncalled \"paths\" satisfying a simple switching property: Whenever two paths P and\nQ intersect, there must be another path that is contained in the beginning of P\nand the end of Q.\n We show that a maximum abstract flow over time can be obtained by solving a\nweighted abstract flow problem and constructing a temporally repeated flow from\nits solution. In the course of the proof, we also show that the relatively\nmodest switching property of abstract networks already captures many essential\nproperties of classical networks.\n", " For given graphs $G$ and $F$, the Tur\\'an number $ex(G,F)$ is defined to be\nthe maximum number of edges in an $F$-free subgraph of $G$. Foucaud,\nKrivelevich and Perarnau and later independently Briggs and Cox introduced a\ndual version of this problem wherein for a given number $k$, one maximizes the\nnumber of edges in a host graph $G$ for which $ex(G,H) < k$.\n Addressing a problem of Briggs and Cox, we determine the asymptotic value of\nthe inverse Tur\\'an number of the paths of length $4$ and $5$ and provide an\nimproved lower bound for all paths of even length. Moreover, we obtain bounds\non the inverse Tur\\'an number of even cycles giving improved bounds on the\nleading coefficient in the case of $C_4$. Finally, we give multiple conjectures\nconcerning the asymptotic value of the inverse Tur\\'an number of $C_4$ and\n$P_{\\ell}$, suggesting that in the latter problem the asymptotic behavior\ndepends heavily on the parity of $\\ell$.\n", " Finding cliques in random graphs and the closely related \"planted\" clique\nvariant, where a clique of size k is planted in a random G(n, 1/2) graph, have\nbeen the focus of substantial study in algorithm design. Despite much effort,\nthe best known polynomial-time algorithms only solve the problem for k ~\nsqrt(n).\n In this paper we study the complexity of the planted clique problem under\nalgorithms from the Sum-of-squares hierarchy. We prove the first average case\nlower bound for this model: for almost all graphs in G(n,1/2), r rounds of the\nSOS hierarchy cannot find a planted k-clique unless k > n^{1/2r} (up to\nlogarithmic factors). Thus, for any constant number of rounds planted cliques\nof size n^{o(1)} cannot be found by this powerful class of algorithms. This is\nshown via an integrability gap for the natural formulation of maximum clique\nproblem on random graphs for SOS and Lasserre hierarchies, which in turn follow\nfrom degree lower bounds for the Positivestellensatz proof system.\n We follow the usual recipe for such proofs. First, we introduce a natural\n\"dual certificate\" (also known as a \"vector-solution\" or \"pseudo-expectation\")\nfor the given system of polynomial equations representing the problem for every\nfixed input graph. Then we show that the matrix associated with this dual\ncertificate is PSD (positive semi-definite) with high probability over the\nchoice of the input graph.This requires the use of certain tools. One is the\ntheory of association schemes, and in particular the eigenspaces and\neigenvalues of the Johnson scheme. Another is a combinatorial method we develop\nto compute (via traces) norm bounds for certain random matrices whose entries\nare highly dependent; we hope this method will be useful elsewhere.\n", " Given $n$ weighted points (positive or negative) in $d$ dimensions, what is\nthe axis-aligned box which maximizes the total weight of the points it\ncontains?\n The best known algorithm for this problem is based on a reduction to a\nrelated problem, the Weighted Depth problem [T. M. Chan, FOCS'13], and runs in\ntime $O(n^d)$. It was conjectured [Barbay et al., CCCG'13] that this runtime is\ntight up to subpolynomial factors. We answer this conjecture affirmatively by\nproviding a matching conditional lower bound. We also provide conditional lower\nbounds for the special case when points are arranged in a grid (a well studied\nproblem known as Maximum Subarray problem) as well as for other related\nproblems.\n All our lower bounds are based on assumptions that the best known algorithms\nfor the All-Pairs Shortest Paths problem (APSP) and for the Max-Weight k-Clique\nproblem in edge-weighted graphs are essentially optimal.\n", " We present results and conjectures on the van der Waerden numbers w(2;3,t)\nand on the new palindromic van der Waerden numbers pdw(2;3,t). We have computed\nthe new number w(2;3,19) = 349, and we provide lower bounds for 20 <= t <= 39,\nwhere for t <= 30 we conjecture these lower bounds to be exact. The lower\nbounds for 24 <= t <= 30 refute the conjecture that w(2;3,t) <= t^2, and we\npresent an improved conjecture. We also investigate regularities in the good\npartitions (certificates) to better understand the lower bounds.\n Motivated by such reglarities, we introduce *palindromic van der Waerden\nnumbers* pdw(k; t_0,...,t_{k-1}), defined as ordinary van der Waerden numbers\nw(k; t_0,...,t_{k-1}), however only allowing palindromic solutions (good\npartitions), defined as reading the same from both ends. Different from the\nsituation for ordinary van der Waerden numbers, these \"numbers\" need actually\nto be pairs of numbers. We compute pdw(2;3,t) for 3 <= t <= 27, and we provide\nlower bounds, which we conjecture to be exact, for t <= 35.\n All computations are based on SAT solving, and we discuss the various\nrelations between SAT solving and Ramsey theory. Especially we introduce a\nnovel (open-source) SAT solver, the tawSolver, which performs best on the SAT\ninstances studied here, and which is actually the original DLL-solver, but with\nan efficient implementation and a modern heuristic typical for look-ahead\nsolvers (applying the theory developed in the SAT handbook article of the\nsecond author).\n", " A result of Boros and F\\\"uredi ($d=2$) and of B\\'ar\\'any (arbitrary $d$)\nasserts that for every $d$ there exists $c_d>0$ such that for every $n$-point\nset $P\\subset \\R^d$, some point of $\\R^d$ is covered by at least $c_d{n\\choose\nd+1}$ of the $d$-simplices spanned by the points of $P$. The largest possible\nvalue of $c_d$ has been the subject of ongoing research. Recently Gromov\nimproved the existing lower bounds considerably by introducing a new,\ntopological proof method. We provide an exposition of the combinatorial\ncomponent of Gromov's approach, in terms accessible to combinatorialists and\ndiscrete geometers, and we investigate the limits of his method. In particular,\nwe give tighter bounds on the \\emph{cofilling profiles} for the\n$(n-1)$-simplex. These bounds yield a minor improvement over Gromov's lower\nbounds on $c_d$ for large $d$, but they also show that the room for further\nimprovement through the {\\cofilling} profiles alone is quite small. We also\nprove a slightly better lower bound for $c_3$ by an approach using an\nadditional structure besides the {\\cofilling} profiles. We formulate a\ncombinatorial extremal problem whose solution might perhaps lead to a tight\nlower bound for $c_d$.\n", " Maximally monotone operators and firmly nonexpansive mappings play key roles\nin modern optimization and nonlinear analysis. Five years ago, it was shown\nthat if finitely many firmly nonexpansive operators are all asymptotically\nregular (i.e., the have or \"almost have\" fixed points), then the same is true\nfor compositions and convex combinations. In this paper, we derive bounds on\nthe magnitude of the minimal displacement vectors of compositions and of convex\ncombinations in terms of the displacement vectors of the underlying operators.\nOur results completely generalize earlier works. Moreover, we present various\nexamples illustrating that our bounds are sharp.\n", " We give upper and lower bounds on the determinant of a perturbation of the\nidentity matrix or, more generally, a perturbation of a nonsingular diagonal\nmatrix. The matrices considered are, in general, diagonally dominant. The lower\nbounds are best possible, and in several cases they are stronger than\nwell-known bounds due to Ostrowski and other authors. If $A = I-E$ is an $n\n\\times n$ matrix and the elements of $E$ are bounded in absolute value by\n$\\varepsilon \\le 1/n$, then a lower bound of Ostrowski (1938) is $\\det(A) \\ge\n1-n\\varepsilon$. We show that if, in addition, the diagonal elements of $E$ are\nzero, then a best-possible lower bound is \\[\\det(A) \\ge\n(1-(n-1)\\varepsilon)\\,(1+\\varepsilon)^{n-1}.\\] Corresponding upper bounds are\nrespectively \\[\\det(A) \\le (1 + 2\\varepsilon + n\\varepsilon^2)^{n/2}\\] and\n\\[\\det(A) \\le (1 + (n-1)\\varepsilon^2)^{n/2}.\\] The first upper bound is\nstronger than Ostrowski's bound (for $\\varepsilon < 1/n$) $\\det(A) \\le (1 -\nn\\varepsilon)^{-1}$. The second upper bound generalises Hadamard's inequality,\nwhich is the case $\\varepsilon = 1$.\n A necessary and sufficient condition for our upper bounds to be best possible\nfor matrices of order $n$ and all positive $\\varepsilon$ is the existence of a\nskew-Hadamard matrix of order $n$.\n", " The entropy of a graph is a functional depending both on the graph itself and\non a probability distribution on its vertex set. This graph functional\noriginated from the problem of source coding in information theory and was\nintroduced by J. K\\\"{o}rner in 1973. Although the notion of graph entropy has\nits roots in information theory, it was proved to be closely related to some\nclassical and frequently studied graph theoretic concepts. For example, it\nprovides an equivalent definition for a graph to be perfect and it can also be\napplied to obtain lower bounds in graph covering problems.\n In this thesis, we review and investigate three equivalent definitions of\ngraph entropy and its basic properties. Minimum entropy colouring of a graph\nwas proposed by N. Alon in 1996. We study minimum entropy colouring and its\nrelation to graph entropy. We also discuss the relationship between the entropy\nand the fractional chromatic number of a graph which was already established in\nthe literature.\n A graph $G$ is called \\emph{symmetric with respect to a functional $F_G(P)$}\ndefined on the set of all the probability distributions on its vertex set if\nthe distribution $P^*$ maximizing $F_G(P)$ is uniform on $V(G)$. Using the\ncombinatorial definition of the entropy of a graph in terms of its vertex\npacking polytope and the relationship between the graph entropy and fractional\nchromatic number, we prove that vertex transitive graphs are symmetric with\nrespect to graph entropy. Furthermore, we show that a bipartite graph is\nsymmetric with respect to graph entropy if and only if it has a perfect\nmatching. As a generalization of this result, we characterize some classes of\nsymmetric perfect graphs with respect to graph entropy. Finally, we prove that\nthe line graph of every bridgeless cubic graph is symmetric with respect to\ngraph entropy.\n", " Classifier-free guidance is a key component for enhancing the performance of\nconditional generative models across diverse tasks. While it has previously\ndemonstrated remarkable improvements for the sample quality, it has only been\nexclusively employed for diffusion models. In this paper, we integrate\nclassifier-free guidance into Flow Matching (FM) models, an alternative\nsimulation-free approach that trains Continuous Normalizing Flows (CNFs) based\non regressing vector fields. We explore the usage of \\emph{Guided Flows} for a\nvariety of downstream applications. We show that Guided Flows significantly\nimproves the sample quality in conditional image generation and zero-shot\ntext-to-speech synthesis, boasting state-of-the-art performance. Notably, we\nare the first to apply flow models for plan generation in the offline\nreinforcement learning setting, showcasing a 10x speedup in computation\ncompared to diffusion models while maintaining comparable performance.\n", " The classical Zarankiewicz's problem asks for the maximum number of edges in\na bipartite graph on $n$ vertices which does not contain the complete bipartite\ngraph $K_{t,t}$. In one of the cornerstones of extremal graph theory,\nK\\H{o}v\\'ari S\\'os and Tur\\'an proved an upper bound of $O(n^{2-\\frac{1}{t}})$.\nIn a celebrated result, Fox et al. obtained an improved bound of\n$O(n^{2-\\frac{1}{d}})$ for graphs of VC-dimension $d$ (where $d<t$). Basit,\nChernikov, Starchenko, Tao and Tran improved the bound for the case of\nsemilinear graphs. At SODA'23, Chan and Har-Peled further improved Basit et\nal.'s bounds and presented (quasi-)linear upper bounds for several classes of\ngeometrically-defined incidence graphs, including a bound of $O(n \\log \\log n)$\nfor the incidence graph of points and pseudo-discs in the plane.\n In this paper we present a new approach to Zarankiewicz's problem, via\n$\\epsilon$-t-nets - a recently introduced generalization of the classical\nnotion of $\\epsilon$-nets. We show that the existence of `small'-sized\n$\\epsilon$-t-nets implies upper bounds for Zarankiewicz's problem. Using the\nnew approach, we obtain a sharp bound of $O(n)$ for the intersection graph of\ntwo families of pseudo-discs, thus both improving and generalizing the result\nof Chan and Har-Peled from incidence graphs to intersection graphs. We also\nobtain a short proof of the $O(n^{2-\\frac{1}{d}})$ bound of Fox et al., and\nshow improved bounds for several other classes of geometric intersection\ngraphs, including a sharp $O(n\\frac{\\log n}{\\log \\log n})$ bound for the\nintersection graph of two families of axis-parallel rectangles.\n", " The aim of this work is to give degree formulas for the generalized Hamming\nweights of evaluation codes and to show lower bounds for these weights. In\nparticular, we give degree formulas for the generalized Hamming weights of\nReed--Muller-type codes, and we determine the minimum distance of toric codes\nover hypersimplices, and the 1st and 2nd generalized Hamming weights of\nsquarefree evaluation codes.\n", " We give here some new lower bounds on the order of a largest induced forest\nin planar graphs with girth $4$ and $5$. In particular we prove that a\ntriangle-free planar graph of order $n$ admits an induced forest of order at\nleast $\\frac{6n+7}{11}$ , improving the lower bound of Salavatipour [M. R.\nSalavatipour, Large induced forests in triangle-free planar graphs, Graphs and\nCombinatorics, 22:113-126, 2006]. We also prove that a planar graph of order\n$n$ and girth at least $5$ admits an induced forest of order at least\n$\\frac{44n+50}{69}$.\n", " We consider straight line drawings of a planar graph $G$ with possible edge\ncrossings. The \\emph{untangling problem} is to eliminate all edge crossings by\nmoving as few vertices as possible to new positions. Let $fix(G)$ denote the\nmaximum number of vertices that can be left fixed in the worst case. In the\n\\emph{allocation problem}, we are given a planar graph $G$ on $n$ vertices\ntogether with an $n$-point set $X$ in the plane and have to draw $G$ without\nedge crossings so that as many vertices as possible are located in $X$. Let\n$fit(G)$ denote the maximum number of points fitting this purpose in the worst\ncase. As $fix(G)\\le fit(G)$, we are interested in upper bounds for the latter\nand lower bounds for the former parameter.\n For each $\\epsilon>0$, we construct an infinite sequence of graphs with\n$fit(G)=O(n^{\\sigma+\\epsilon})$, where $\\sigma<0.99$ is a known graph-theoretic\nconstant, namely the shortness exponent for the class of cubic polyhedral\ngraphs. To the best of our knowledge, this is the first example of graphs with\n$fit(G)=o(n)$. On the other hand, we prove that $fix(G)\\ge\\sqrt{n/30}$ for all\n$G$ with tree-width at most 2. This extends the lower bound obtained by Goaoc\net al. [Discrete and Computational Geometry 42:542-569 (2009)] for outerplanar\ngraphs.\n Our upper bound for $fit(G)$ is based on the fact that the constructed graphs\ncan have only few collinear vertices in any crossing-free drawing. To prove the\nlower bound for $fix(G)$, we show that graphs of tree-width 2 admit drawings\nthat have large sets of collinear vertices with some additional special\nproperties.\n", " Over the years, scene understanding has attracted a growing interest in\ncomputer vision, providing the semantic and physical scene information\nnecessary for robots to complete some particular tasks autonomously. In 3D\nscenes, rich spatial geometric and topological information are often ignored by\nRGB-based approaches for scene understanding. In this study, we develop a\nbottom-up approach for scene understanding that infers support relations\nbetween objects from a point cloud. Our approach utilizes the spatial topology\ninformation of the plane pairs in the scene, consisting of three major steps.\n1) Detection of pairwise spatial configuration: dividing primitive pairs into\nlocal support connection and local inner connection; 2) primitive\nclassification: a combinatorial optimization method applied to classify\nprimitives; and 3) support relations inference and hierarchy graph\nconstruction: bottom-up support relations inference and scene hierarchy graph\nconstruction containing primitive level and object level. Through experiments,\nwe demonstrate that the algorithm achieves excellent performance in primitive\nclassification and support relations inference. Additionally, we show that the\nscene hierarchy graph contains rich geometric and topological information of\nobjects, and it possesses great scalability for scene understanding.\n", " Solutions to linear controlled differential equations can be expressed in\nterms of iterated path integrals of the driving path. This collection of\niterated integrals encodes essentially all information about the driving path.\nWhile upper bounds for iterated path integrals are well known, lower bounds are\nmuch less understood, and it is known only relatively recently that some type\nof asymptotics for the $n$-th order iterated integral can be used to recover\nsome intrinsic quantitative properties of the path, such as the length of $C^1$\npaths.\n In the present paper, we investigate the simplest type of rough paths (the\nrough path analogue of line segments), and establish uniform upper and lower\nestimates for the tail asymptotics of iterated integrals in terms of the local\nvariation of the path. Our methodology, which we believe is new for this\nproblem, involves developing paths into complex semisimple Lie algebras and\nusing the associated representation theory to study spectral properties of Lie\npolynomials under the Lie algebraic development.\n", " We study the Art Gallery Problem for face guards in polyhedral environments.\nThe problem can be informally stated as: how many (not necessarily convex)\nwindows should we place on the external walls of a dark building, in order to\ncompletely illuminate its interior?\n We consider both closed and open face guards (i.e., faces with or without\ntheir boundary), and we study several classes of polyhedra, including\northogonal polyhedra, 4-oriented polyhedra, and 2-reflex orthostacks.\n We give upper and lower bounds on the minimum number of faces required to\nguard the interior of a given polyhedron in each of these classes, in terms of\nthe total number of its faces, f. In several cases our bounds are tight: f/6\nopen face guards for orthogonal polyhedra and 2-reflex orthostacks, and f/4\nopen face guards for 4-oriented polyhedra. Additionally, for closed face guards\nin 2-reflex orthostacks, we give a lower bound of (f+3)/9 and an upper bound of\n(f+1)/7.\n Then we show that it is NP-hard to approximate the minimum number of (closed\nor open) face guards within a factor of Omega(log f), even for polyhedra that\nare orthogonal and simply connected. We also obtain the same hardness results\nfor polyhedral terrains.\n Along the way we discuss some applications, arguing that face guards are not\na reasonable model for guards patrolling on the surface of a polyhedron.\n", " We study nonlinear degenerate parabolic equations of Fokker-Planck type which\ncan be viewed as gradient flows with respect to the recently introduced\nspherical Hellinger-Kantorovich distance. The driving entropy is not assumed to\nbe geodesically convex. We prove solvability of the problem and the\nentropy-entropy production inequality, which implies exponential convergence to\nthe equilibrium. As a corollary, we obtain some related results for the\nWasserstein gradient flows. We also deduce transportation inequalities in the\nspirit of Talagrand, Otto and Villani for the spherical and conic\nHellinger-Kantorovich distances.\n", " In this paper we determine the graph whose least eigenvalue of signless\nLaplacian attains the minimum or maximum among all connected non-bipartite\ngraphs of fixed order and given number of pendant vertices. Thus we obtain a\nlower bound and an upper bound for the least eigenvalue of signless Laplacian\nof a graph in terms of the number of pendent vertices.\n", " The k-means algorithm is a well-known method for partitioning n points that\nlie in the d-dimensional space into k clusters. Its main features are\nsimplicity and speed in practice. Theoretically, however, the best known upper\nbound on its running time (i.e. O(n^{kd})) can be exponential in the number of\npoints. Recently, Arthur and Vassilvitskii [3] showed a super-polynomial\nworst-case analysis, improving the best known lower bound from \\Omega(n) to\n2^{\\Omega(\\sqrt{n})} with a construction in d=\\Omega(\\sqrt{n}) dimensions. In\n[3] they also conjectured the existence of superpolynomial lower bounds for any\nd >= 2.\n Our contribution is twofold: we prove this conjecture and we improve the\nlower bound, by presenting a simple construction in the plane that leads to the\nexponential lower bound 2^{\\Omega(n)}.\n", " Graph Visualization, also known as Graph Drawing, aims to find geometric\nembeddings of graphs that optimize certain criteria. Stress is a widely used\nmetric; stress is minimized when every pair of nodes is positioned at their\nshortest path distance. However, stress optimization presents computational\nchallenges due to its inherent complexity and is usually solved using\nheuristics in practice. We introduce a scalable Graph Neural Network (GNN)\nbased Graph Drawing framework with sub-quadratic runtime that can learn to\noptimize stress. Inspired by classical stress optimization techniques and\nforce-directed layout algorithms, we create a coarsening hierarchy for the\ninput graph. Beginning at the coarsest level, we iteratively refine and\nun-coarsen the layout, until we generate an embedding for the original graph.\nTo enhance information propagation within the network, we propose a novel\npositional rewiring technique based on intermediate node positions. Our\nempirical evaluation demonstrates that the framework achieves state-of-the-art\nperformance while remaining scalable.\n", " Based on earlier work by Carlen-Maas and the second- and third-named author,\nwe introduce the notion of intertwining curvature lower bounds for graphs and\nquantum Markov semigroups. This curvature notion is stronger than both\nBakry-\\'Emery and entropic Ricci curvature, while also computationally simpler\nthan the latter. We verify intertwining curvature bounds in a number of\nexamples, including finite weighted graphs and graphs with Laplacians admitting\nnice mapping representations, as well as generalized dephasing semigroups and\nquantum Markov semigroups whose generators are formed by commuting jump\noperators. By improving on the best-known bounds for entropic curvature of\ndepolarizing semigroups, we demonstrate that there can be a gap between the\noptimal intertwining and entropic curvature bound. In the case of qubits, this\nimproved entropic curvature bound implies the modified logarithmic Sobolev\ninequality with optimal constant.\n", " The Frechet distance is a well-studied and very popular measure of similarity\nof two curves. Many variants and extensions have been studied since Alt and\nGodau introduced this measure to computational geometry in 1991. Their original\nalgorithm to compute the Frechet distance of two polygonal curves with n\nvertices has a runtime of O(n^2 log n). More than 20 years later, the state of\nthe art algorithms for most variants still take time more than O(n^2 / log n),\nbut no matching lower bounds are known, not even under reasonable complexity\ntheoretic assumptions.\n To obtain a conditional lower bound, in this paper we assume the Strong\nExponential Time Hypothesis or, more precisely, that there is no\nO*((2-delta)^N) algorithm for CNF-SAT for any delta > 0. Under this assumption\nwe show that the Frechet distance cannot be computed in strongly subquadratic\ntime, i.e., in time O(n^{2-delta}) for any delta > 0. This means that finding\nfaster algorithms for the Frechet distance is as hard as finding faster CNF-SAT\nalgorithms, and the existence of a strongly subquadratic algorithm can be\nconsidered unlikely.\n Our result holds for both the continuous and the discrete Frechet distance.\nWe extend the main result in various directions. Based on the same assumption\nwe (1) show non-existence of a strongly subquadratic 1.001-approximation, (2)\npresent tight lower bounds in case the numbers of vertices of the two curves\nare imbalanced, and (3) examine realistic input assumptions (c-packed curves).\n", " In this paper our aim is to prove some monotonicity and convexity results for\nthe modified Struve function of the second kind by using its integral\nrepresentation. Moreover, as consequences of these results, we present some\nfunctional inequalities (like Tur\\'an type inequalities) as well as lower and\nupper bounds for modified Struve function of the second kind and its\nlogarithmic derivative.\n", " We derive a dyadic model operator for the Riesz vector. We show linear lower\n$L^p$ bounds for $1 < p < \\infty$ between this model operator and the Riesz\nvector, when applied to functions with values in Banach spaces. By a lower\nbound we mean that the boundedness of the Riesz vector implies the boundedness\nof the dydic Riesz vector.\n", " It is known from the algebraic graph theory that if $L$ is the Laplacian\nmatrix of some tree $G$ with a vertex degree sequence $\\mathbf{d}=(d_1, ...,\nd_n)^\\top$ and $D$ is its distance matrix, then\n$LD+2I=(2\\cdot\\mathbf{1}-\\mathbf{d})\\mathbf{1}^\\top$, where $\\mathbf{1}$ is an\nall-ones column vector. We prove that if this matrix identity holds for the\nLaplacian matrix of some graph $G$ with a degree sequence $\\mathbf{d}$ and for\nsome matrix $D$, then $G$ is essentially a tree, and $D$ is its distance\nmatrix. This result immediately generalizes to weighted graphs. If the matrix\n$D$ is symmetric, the lower triangular part of this matrix identity is\nredundant and can be omitted. Therefore, the above bilinear matrix equation in\n$L$, $D$, and $\\mathbf{d}$ characterizes trees in terms of their Laplacian and\ndistance matrices. Applications to the extremal graph theory (especially, to\ntopological index optimization and to optimal tree problems) and to road\ntopology design are discussed.\n", " Bilevel Optimization Programming is used to model complex and conflicting\ninteractions between agents, for example in Robust AI or Privacy-preserving AI.\nIntegrating bilevel mathematical programming within deep learning is thus an\nessential objective for the Machine Learning community. Previously proposed\napproaches only consider single-level programming. In this paper, we extend\nexisting single-level optimization programming approaches and thus propose\nDifferentiating through Bilevel Optimization Programming (BiGrad) for\nend-to-end learning of models that use Bilevel Programming as a layer. BiGrad\nhas wide applicability and can be used in modern machine learning frameworks.\nBiGrad is applicable to both continuous and combinatorial Bilevel optimization\nproblems. We describe a class of gradient estimators for the combinatorial case\nwhich reduces the requirements in terms of computation complexity; for the case\nof the continuous variable, the gradient computation takes advantage of the\npush-back approach (i.e. vector-jacobian product) for an efficient\nimplementation. Experiments show that the BiGrad successfully extends existing\nsingle-level approaches to Bilevel Programming.\n" ]

[ " Molecular Communication (MC) architectures suffer from molecular build-up in\nthe channel if they do not have appropriate reuptake mechanisms. The molecular\nbuild-up either leads to intersymbol interference (ISI) or reduces the\ntransmission rate. To measure the molecular build-up, we derive analytic\nexpressions for the incidence rate and absorption rate for one-dimensional MC\nchannels where molecular dispersion obeys the Brownian Motion. We verify each\nof our key results with Monte Carlo simulations. Our results contribute to the\ndevelopment of more complicated models and analytic expressions to measure the\nmolecular build-up and the impact of ISI in MC.\n", " Polymer Optical Fiber (POF) can be and are being used in various fields of\napplications. Two of the main fields are the automotive and the home\nentertainment sector. The POF can be applied in several different optical\ncommunication systems as automotive multi-media busses or in-house Ethernet\nsystems.\n The requirements of bandwidth are increasing very fast in these sectors and\ntherefore solutions that satisfy these demands are of high actuality. One\nsolution is to use the wavelength division multiplexing (WDM) technique. Here,\nseveral different wavelengths can carry information over one POF fiber. All\nwavelengths that are transmitted over the fiber, must be separated at the\nreceiver to regain and redirect the information channels. These separators are\nso-called Demultiplexers.\n There are several systems available on the market, which are all afflicted\nwith certain disadvantages. But all these solutions have one main disadvantage,\nthey are all too expensive for most of the applications mentioned above. So the\ngoal of this study is to develop an economical Demultiplexer for WDM\ntransmission over POF.\n The main idea is to separate the chromatic light in its monochromatic\ncomponents with the help of a prism with low reciprocal dispersive power. The\nprism and the other assemblies, which are needed to adjust the optical path,\nshould be manufactured in injection molding technique. This manufacturing\ntechnique is a very simple and economical way to produce a mass production\napplicable Demultiplexer for POF.\n", " Semantic communication allows the receiver to know the intention instead of\nthe bit information itself, which is an emerging technique to support real-time\nhuman-machine and machine-to-machine interactions for future wireless\ncommunications. In semantic communications, both transmitter and receiver share\nsome common knowledge, which can be used to extract small-size information at\nthe transmitter and recover the original information at the receiver. Due to\ndifferent design purposes, security issues in semantic communications have two\nunique features compared to standard bit-wise communications. First, an\nattacker in semantic communications considers not only the amount of stolen\ndata but also the meanings of stolen data. Second, an attacker in semantic\ncommunication systems can attack not only semantic information transmission as\ndone in standard communication systems but also attacks machine learning (ML)\nmodels used for semantic information extraction since most of semantic\ninformation is generated using ML based methods. Due to these unique features,\nin this paper, we present an overview on the fundamentals and key challenges in\nthe design of secure semantic communication. We first provide various methods\nto define and extract semantic information. Then, we focus on secure semantic\ncommunication techniques in two areas: information security and semantic ML\nmodel security. For each area, we identify the main problems and challenges.\nThen, we will provide a comprehensive treatment of these problems. In a\nnutshell,this article provides a holistic set of guidelines on how to design\nsecure semantic communication systems over real-world wireless communication\nnetworks.\n", " In this paper, we introduce the notion of Plausible Deniability in an\ninformation theoretic framework. We consider a scenario where an entity that\neavesdrops through a broadcast channel summons one of the parties in a\ncommunication protocol to reveal their message (or signal vector). It is\ndesirable that the summoned party have enough freedom to produce a fake output\nthat is likely plausible given the eavesdropper's observation. We examine three\nvariants of this problem -- Message Deniability, Transmitter Deniability, and\nReceiver Deniability. In the first setting, the message sender is summoned to\nproduce the sent message. Similarly, in the second and third settings, the\ntransmitter and the receiver are required to produce the transmitted codeword,\nand the received vector respectively. For each of these settings, we examine\nthe maximum communication rate that allows a given minimum rate of plausible\nfake outputs. For the Message and Transmitter Deniability problems, we fully\ncharacterise the capacity region for general broadcast channels, while for the\nReceiver Deniability problem, we give an achievable rate region for physically\ndegraded broadcast channels.\n", " First, we show a new inexpensive defense against intruders and the\nman-in-the-middle attack in the Kirchhoff's-loop-Johnson-like-noise (KLJN)\ncipher. Then instead of point-to-point communication, we propose a high\nefficiency, secure network. The (in the idealistic case totally secure)\nclassical network is based on an improved version of the KLJN cipher. The\nnetwork consists of two parallel networks: i) a chain-like network of securely\ncommunicating, electrically isolated Kirchhoff-loops with Johnson-like noise\nand driven by a specific switching process of the resistances; ii) and a\nregular non-secure data network with a Coordinator-server. If the classical\nnetwork is fast enough, the chain-like network of N communicators can generate\nand share an N bit long secret key within a single clock period of the ciphers\nand that implies a significant speed-up compared to the point-to-point key\nexchanges used by quantum communication or RSA-like key exchange methods. This\nis a teleportation-type multiple telecloning of the classical information bit\nbecause the information transfer can take place without the actual presence of\nthe information bit at the intermediate points of the network. With similar\nquantum schemes the telecloning of classical bits via quantum communicator\nnetworks without telecloning the quantum states is also possible.\n", " We show that quantum communication by means of collapse of the wave function\nis possible. In this study, quantum communication does not mean quantum\nteleportation or quantum cryptography, but transmission of information itself.\nBecause of consistency with special relativity, the possibility of the quantum\ncommunication leads to another conclusion that the collapse of the wave\nfunction must propagate at the speed of light or slower.\n We show this requirement is consistent with nonlocality in quantum mechanics.\nWe also demonstrate that the Einstein-Podolsky-Rosen experiment does not\ndisprove our conclusion.\n", " Recent interest in integrated sensing and communications has led to the\ndesign of novel signal processing techniques to recover information from an\noverlaid radar-communications signal. Here, we focus on a spectral coexistence\nscenario, wherein the channels and transmit signals of both radar and\ncommunications systems are unknown to the common receiver. In this dual-blind\ndeconvolution (DBD) problem, the receiver admits a multi-carrier wireless\ncommunications signal that is overlaid with the radar signal reflected off\nmultiple targets. The communications and radar channels are represented by\ncontinuous-valued range-times or delays corresponding to multiple transmission\npaths and targets, respectively. Prior works addressed recovery of unknown\nchannels and signals in this ill-posed DBD problem through atomic norm\nminimization but contingent on individual minimum separation conditions for\nradar and communications channels. In this paper, we provide an optimal joint\nseparation condition using extremal functions from the Beurling-Selberg\ninterpolation theory. Thereafter, we formulate DBD as a low-rank modified\nHankel matrix retrieval and solve it via nuclear norm minimization. We estimate\nthe unknown target and communications parameters from the recovered low-rank\nmatrix using multiple signal classification (MUSIC) method. We show that the\njoint separation condition also guarantees that the underlying Vandermonde\nmatrix for MUSIC is well-conditioned. Numerical experiments validate our\ntheoretical findings.\n", " We propose a targeted communication architecture for multi-agent\nreinforcement learning, where agents learn both what messages to send and whom\nto address them to while performing cooperative tasks in partially-observable\nenvironments. This targeting behavior is learnt solely from downstream\ntask-specific reward without any communication supervision. We additionally\naugment this with a multi-round communication approach where agents coordinate\nvia multiple rounds of communication before taking actions in the environment.\nWe evaluate our approach on a diverse set of cooperative multi-agent tasks, of\nvarying difficulties, with varying number of agents, in a variety of\nenvironments ranging from 2D grid layouts of shapes and simulated traffic\njunctions to 3D indoor environments, and demonstrate the benefits of targeted\nand multi-round communication. Moreover, we show that the targeted\ncommunication strategies learned by agents are interpretable and intuitive.\nFinally, we show that our architecture can be easily extended to mixed and\ncompetitive environments, leading to improved performance and sample complexity\nover recent state-of-the-art approaches.\n", " For this year's 8TeV run of the LHC we lay out different strategies to search\nfor scalar top pairs. We show results for the hadronic and for the\nsemi-leptonic channels based on hadronic top tagging. For the di-lepton channel\nwe illustrate the impact of transverse mass variables. Each of our\nsignal-to-background ratios ranges around unity for a stop mass around 400GeV.\nThe combined signal significances show that dedicated stop searches are\nbecoming sensitive over a non-negligible part of parameter space.\n", " The present study focuses on the interaction of gravity waves in the\natmosphere with the tropopause. As the vertical extent of the latter is small\ncompared to the density scale height, wave propagation is described by the\nTaylor-Goldstein equation as derived from the linearised Boussinesq\napproximation. Of particular interest in the construction of gravity wave\nparameterisations for the upper atmosphere are the transmission and reflection\nproperties of the tropopause as these determine the upward fluxes of energy and\nmomentum carried by internal waves.\n A method is presented that decomposes internal waves explicitly into upward\nand downward propagating contributions, thus giving direct access to\ntransmission and reflection coefficients of finite regions of non-uniform\nstratification in a stationary atmosphere. The scheme utilizes a piecewise\nconstant approximation for the background stratification and matches up- and\ndownward propagating plane wave solutions in each layer through physically\nmeaningful coupling conditions. As a result, transmission and reflection\ncoefficients follow immediately.\n", " We present a model for neutrino-nucleus scattering in the energy region\nrelevant for present and forthcoming neutrino-oscillation experiments. The\nmodel is based on the RPA treatment of the nuclear responses in the\nquasi-elastic and Delta-resonance region. It includes also in a\nphenomenological way nucleon knock-out. It aims at the description, within a\nsingle framework, of several final state channels i.e. quasi-elastic,\nincoherent and coherent one-pion production and two- or several-nucleon\nknock-out.\n", " Traffic signal control (TSC) is a challenging problem within intelligent\ntransportation systems and has been tackled using multi-agent reinforcement\nlearning (MARL). While centralized approaches are often infeasible for\nlarge-scale TSC problems, decentralized approaches provide scalability but\nintroduce new challenges, such as partial observability. Communication plays a\ncritical role in decentralized MARL, as agents must learn to exchange\ninformation using messages to better understand the system and achieve\neffective coordination. Deep MARL has been used to enable inter-agent\ncommunication by learning communication protocols in a differentiable manner.\nHowever, many deep MARL communication frameworks proposed for TSC allow agents\nto communicate with all other agents at all times, which can add to the\nexisting noise in the system and degrade overall performance. In this study, we\npropose a communication-based MARL framework for large-scale TSC. Our framework\nallows each agent to learn a communication policy that dictates \"which\" part of\nthe message is sent \"to whom\". In essence, our framework enables agents to\nselectively choose the recipients of their messages and exchange variable\nlength messages with them. This results in a decentralized and flexible\ncommunication mechanism in which agents can effectively use the communication\nchannel only when necessary. We designed two networks, a synthetic $4 \\times 4$\ngrid network and a real-world network based on the Pasubio neighborhood in\nBologna. Our framework achieved the lowest network congestion compared to\nrelated methods, with agents utilizing $\\sim 47-65 \\%$ of the communication\nchannel. Ablation studies further demonstrated the effectiveness of the\ncommunication policies learned within our framework.\n", " We consider a mobile edge computing scenario where users want to perform a\nlinear inference operation $\\boldsymbol{W} \\boldsymbol{x}$ on local data\n$\\boldsymbol{x}$ for some network-side matrix $\\boldsymbol{W}$. The inference\nis performed in a distributed fashion over multiple servers at the network\nedge. For this scenario, we propose a coding scheme that combines a rateless\ncode to provide resiliency against straggling servers--hence reducing the\ncomputation latency--and an irregular-repetition code to provide spatial\ndiversity--hence reducing the communication latency. We further derive a lower\nbound on the total latency--comprising computation latency, communication\nlatency, and decoding latency. The proposed scheme performs remarkably close to\nthe bound and yields significantly lower latency than the scheme based on\nmaximum distance separable codes recently proposed by Zhang and Simeone.\n", " We investigate the secure connectivity of wireless sensor networks under the\npairwise key distribution scheme of Chan et al.. Unlike recent work which was\ncarried out under the assumption of full visibility, here we assume a\n(simplified) communication model where unreliable wireless links are\nrepresented as on/off channels. We present conditions on how to scale the model\nparameters so that the network i) has no secure node which is isolated and ii)\nis securely connected, both with high probability when the number of sensor\nnodes becomes large. The results are given in the form of zero-one laws, and\nexhibit significant differences with corresponding results in the full\nvisibility case. Through simulations these zero-one laws are shown to be valid\nalso under a more realistic communication model, i.e., the disk model.\n", " In the past, a maximum-entropy model was introduced and applied to the study\nof statistical scattering by chaotic cavities, when short paths may play an\nimportant role in the scattering process. In particular, the validity of the\nmodel was investigated in relation with the statistical properties of the\nconductance in open chaotic cavities. In this article we investigate further\nthe validity of the maximum-entropy model, by comparing the theoretical\npredictions with the results of computer simulations, in which the Schroedinger\nequation is solved numerically inside the cavity for one and two open channels\nin the leads; we analyze, in addition to the conductance, the zero-frequency\nlimit of the shot-noise power spectrum. We also obtain theoretical results for\nthe ensemble average of this last quantity, for the orthogonal and unitary\ncases of the circular ensemble and an arbitrary number of channels. Generally\nspeaking, the agreement between theory and numerics is good. In some of the\ncavities that we study, short paths consist of whispering gallery modes, which\nwere excluded in previous studies. These cavities turn out to be all the more\ninteresting, as it is in relation with them that we found certain systematic\ndiscrepancies in the comparison with theory. We give evidence that it is the\nlack of stationarity inside the energy interval that is analyzed, and hence the\nlack of ergodicity that gives rise to the discrepancies. Indeed, the agreement\nbetween theory and numerical simulations is improved when the energy interval\nis reduced to a point and the statistics is then collected over an ensemble. It\nthus appears that the maximum-entropy model is valid beyond the domain where it\nwas originally derived. An understanding of this situation is still lacking at\nthe present moment.\n", " General Principle of Relativity unequivocally supports the notion of\nmomentum-less energy for bodies (energy-quanta) moving at the {\\em same} or\n{\\em constant} speed relative to all the reference systems. In this\ncommunication, we point out that whether energy-quantum is a momentum-less body\nor not is verifiable using ultra-cold atoms trapped in an optical lattice,\nperhaps with some minor modifications to the existing such experimental setups.\n", " Directed transport is a key concept for many ongoing applications including\nnanoscale heat management, current rectification, source protection, and energy\nharvesting. Within the context of quantum transport, we here explore the use of\nnonlinear effects introduced by current-induced forces (CIFs) as a practical\nway to effectively break charge and heat transport reciprocities. In\nparticular, we consider a simple model consisting of a mobile quantum dot (QD)\ncoupled to two leads, where the charge (or heat) current develops an asymmetric\nbehavior under inversion of voltage (or temperature) bias, thereby turning the\nsystem into a quantum diode (or quantum thermal diode). Furthermore, we find\nmultiple stable positions for the QD and we show how the extraction of useful\nwork is possible by modulating the nonequilibrium sources along\nwell-established hysteresis loops. Finally, we explore a particular case where\nthe nonlinearity of the CIFs can be exploited to pump heat or charge, even for\nsystems that preserve inversion symmetry. This counterintuitive result is\nattributed to a spontaneous breaking of the inversion symmetry due to the\nintrinsic system's dynamics.\n", " Consider an n-point metric M = (V,delta), and a transmission range assignment\nr: V \\rightarrow \\mathbb R^+ that maps each point v in V to the disk of radius\nr(v) around it. The {symmetric disk graph} (henceforth, SDG) that corresponds\nto M and r is the undirected graph over V whose edge set includes an edge (u,v)\nif both r(u) and r(v) are no smaller than delta(u,v). SDGs are often used to\nmodel wireless communication networks.\n Abu-Affash, Aschner, Carmi and Katz (SWAT 2010, \\cite{AACK10}) showed that\nfor any {2-dimensional Euclidean} n-point metric M, the weight of the MST of\nevery {connected} SDG for M is O(log n) w(MST(M)), and that this bound is\ntight. However, the upper bound proof of \\cite{AACK10} relies heavily on basic\ngeometric properties of 2-dimensional Euclidean metrics, and does not extend to\nhigher dimensions. A natural question that arises is whether this surprising\nupper bound of \\cite{AACK10} can be generalized for wider families of metrics,\nsuch as 3-dimensional Euclidean metrics.\n In this paper we generalize the upper bound of Abu-Affash et al.\n\\cite{AACK10} for Euclidean metrics of any dimension. Furthermore, our upper\nbound extends to {arbitrary metrics} and, in particular, it applies to any of\nthe normed spaces ell_p. Specifically, we demonstrate that for {any} n-point\nmetric M, the weight of the MST of every connected SDG for M is O(log n)\nw(MST(M)).\n", "4z(Assadi, Chen, and Khanna, 2019) define a 4-player hidden-pointer-chasing\n($\\mathsf{HPC}^4$), and using it, give strong multi-pass lower bounds for graph\nproblems in the streaming model of computation and a lower bound on the query\ncomplexity of sub-modular minimization. We present a two-player version\n($\\mathsf{HPC}^2$) of $\\mathsf{HPC}^4$ that has matching communication\ncomplexity to $\\mathsf{HPC}^4$. Our formulation allows us to lower bound its\ncommunication complexity with a simple direct-sum argument. Using this lower\nbound on the communication complexity of $\\mathsf{HPC}^2$, we retain the\nstreaming and query complexity lower bounds by (Assadi, Chen, and Khanna,\n2019).\n Further, by giving reductions from $\\mathsf{HPC}^2$, we prove new multi-pass\nspace lower bounds for graph problems in turnstile streams. In particular, we\nshow that any algorithm which computes the exact weight of the maximum weighted\nmatching in an $n$-vertex graph requires $\\tilde{O}(n^{2})$ space unless it\nmakes $\\omega(\\log n)$ passes over the turnstile stream, and that any algorithm\nwhich computes the minimum $s\\text{-}t$ distance in an $n$-vertex graph\nrequires $n^{2-o(1)}$ space unless it makes $n^{\\Omega(1)}$ passes over the\nturnstile stream. Our reductions can be modified to use $\\mathsf{HPC}^4$ as\nwell.\n", " A multi-layered structure consisting of alternate right-handed material (RHM)\nand left-handed material (LHM)is considered and the unusual narrow transmission\nbands are explained as the competitive results of the Bragg condition and the\ntransparent condition. These unusual narrow transmission bands may exist\nregardless whether the optical length of the LHM layer cancels the optical\nlength of the RHM layer or not. This unusual transmission property may\ndisappear when the reflection coefficient for each interface is small and the\noptical length of the LHM layer does not cancel the optical length of the RHM\nlayer. An non-ideal model when the LHM is dispersive and lossy is also employed\nto confirm the unusual transmission phenomenon.\n", " Communication is a critical factor for the big multi-agent world to stay\norganized and productive. Typically, most previous multi-agent\n\"learning-to-communicate\" studies try to predefine the communication protocols\nor use technologies such as tabular reinforcement learning and evolutionary\nalgorithm, which can not generalize to changing environment or large collection\nof agents.\n In this paper, we propose an Actor-Coordinator-Critic Net (ACCNet) framework\nfor solving \"learning-to-communicate\" problem. The ACCNet naturally combines\nthe powerful actor-critic reinforcement learning technology with deep learning\ntechnology. It can efficiently learn the communication protocols even from\nscratch under partially observable environment. We demonstrate that the ACCNet\ncan achieve better results than several baselines under both continuous and\ndiscrete action space environments. We also analyse the learned protocols and\ndiscuss some design considerations.\n", " We investigate the reflection and refraction behaviors of electromagnetic\nwaves at the interface between an isotropic material and the anisotropic medium\nwith a unique dispersion relation. We show that the refraction angle of whether\nphase or energy flow for $E$-polarized waves is opposite to that for\n$H$-polarized waves, though the dispersion relations for $E$- and $H$-polarized\nwaves are the same in such anisotropic media. For a certain polarized wave the\nrefraction behaviors of wave vector and energy flow are also significantly\ndifferent. It is found that waves exhibit different propagation behaviors in\nanisotropic media with different sign combinations of the permittivity and\npermeability tensors. Some interesting properties of propagation are also found\nin the special anisotropic media, such as amphoteric refraction, oblique total\ntransmission for both $E$- and $H$-polarized waves, and the inversion of\ncritical angle. The special anisotropic media can be realized by metamaterials\nand lead to potential applications, such as to fabricate polarization splitters\nwith higher efficiencies than conventional counterparts.\n", " Quantitatively predictive models of biomolecular circuits are important tools\nfor the design of synthetic biology and molecular communication circuits. The\ninformation content of typical time-lapse single-cell data for the inference of\nkinetic parameters is not only limited by measurement uncertainty and intrinsic\nstochasticity, but also by the employed perturbations. Novel microfluidic\ndevices enable the synthesis of temporal chemical concentration profiles. The\ninformativeness of a perturbation can be quantified based on mutual\ninformation. We propose an approximate method to perform optimal experimental\ndesign of such perturbation profiles. To estimate the mutual information we\nperform a multivariate log-normal approximation of the joint distribution over\nparameters and observations and scan the design space using Metropolis-Hastings\nsampling. The method is demonstrated by finding optimal perturbation sequences\nfor synthetic case studies on a gene expression model with varying reporter\ncharacteristics.\n", " Author offers and researches a new idea - filling tubes by electronic gases.\nHe shows: If the insulating envelope (cover) of the tube is charged positively,\nthe electrons within the tube are not attracted to covering. Tube (as a whole)\nremains a neutral (uncharged) body. The electron gas in the tube has very low\ndensity and very high conductivity, close to superconductivity. If we take the\ndensity (pressure) of electron gas as equal to atmospheric pressure, the\nthickness of insulator film may be very small and the resulting tube is very\nlight.\n Author shows the offered tubes can be applied to many technical fields. For\nexample: (1) Transfer of energy over very long distance with very small\nelectric losses. (2) Design of cheap high altitude electric lines without\nmasts. (3) Transfer of energy from one continent to another continent through\nthe ionosphere. (4) Transfer of a plasma beam (which can convey thrust and\nenergy) from Earth surface to a space ship. (5) Observation of the sky by\ntelescope without atmospheric hindrances. (6) Dirigibles (air balloons) of the\nhighest lift force. (7) Increasing of gun range severalfold. (8) Transfer of\nmatter. And so on.\n Key words: AB tubes, electronic tubes, superconductivity, transmission\nenergy.\n", " Analog beamforming is an attractive and cost-effective solution to exploit\nthe benefits of massive multiple-input-multiple-output systems, by requiring\nonly one up/down-conversion chain. However, the presence of only one chain\nimposes a significant overhead in estimating the channel state information\nrequired for beamforming, when conventional digital channel estimation (CE)\napproaches are used. As an alternative, this paper proposes a novel CE\ntechnique, called periodic analog CE (PACE), that can be performed by analog\nhardware. By avoiding digital processing, the estimation overhead is\nsignificantly lowered and does not scale with number of antennas. PACE involves\nperiodic transmission of a sinusoidal reference signal by the transmitter,\nestimation of its amplitude and phase at each receive antenna via analog\nhardware, and using these estimates for beamforming. To enable such non-trivial\noperation, two reference tone recovery techniques and a novel receiver\narchitecture for PACE are proposed and analyzed, both theoretically and via\nsimulations. Results suggest that in sparse, wide-band channels and above a\ncertain signal-to-noise ratio, PACE aided beamforming suffers only a small loss\nin beamforming gain and enjoys a much lower CE overhead, in comparison to\nconventional approaches. Benefits of using PACE aided beamforming during the\ninitial access phase are also discussed.\n", " When the users in a MIMO broadcast channel experience different spatial\ntransmit correlation matrices, a class of gains is produced that is denoted\ntransmit correlation diversity. This idea was conceived for channels in which\ntransmit correlation matrices have mutually exclusive eigenspaces, allowing\nnon-interfering training and transmission. This paper broadens the scope of\ntransmit correlation diversity to the case of partially and fully overlapping\neigenspaces and introduces techniques to harvest these generalized gains. For\nthe two-user MIMO broadcast channel, we derive achievable degrees of freedom\n(DoF) and achievable rate regions with/without channel state information at the\nreceiver (CSIR). When CSIR is available, the proposed achievable DoF region is\ntight in some configurations of the number of receive antennas and the channel\ncorrelation ranks. We then extend the DoF results to the $K$-user case by\nanalyzing the interference graph that characterizes the overlapping structure\nof the eigenspaces. Our achievability results employ a combination of product\nsuperposition in the common part of the eigenspaces, and pre-beamforming (rate\nsplitting) to create multiple data streams in non-overlapping parts of the\neigenspaces. Massive MIMO is a natural example in which spatially correlated\nlink gains are likely to occur. We study the achievable downlink sum rate for a\nfrequency-division duplex massive MIMO system under transmit correlation\ndiversity.\n", " Vertical federated learning (VFL), where data features are stored in multiple\nparties distributively, is an important area in machine learning. However, the\ncommunication complexity for VFL is typically very high. In this paper, we\npropose a unified framework by constructing coresets in a distributed fashion\nfor communication-efficient VFL. We study two important learning tasks in the\nVFL setting: regularized linear regression and $k$-means clustering, and apply\nour coreset framework to both problems. We theoretically show that using\ncoresets can drastically alleviate the communication complexity, while nearly\nmaintain the solution quality. Numerical experiments are conducted to\ncorroborate our theoretical findings.\n", " We introduce the dune-curvilineargrid module. The module provides the\nself-contained, parallel grid manager, as well as the underlying elementary\ncurvilinear geometry module dune-curvilineargeometry. This work is motivated by\nthe need for reliable and scalable electromagnetic design of nanooptical\ndevices. Curvilinear geometries improve both the accuracy of modeling smooth\nmaterial boundaries, and the h/p-convergence rate of PDE solutions, reducing\nthe necessary computational effort. dune-curvilineargrid provides a large\nspectrum of features for scalable parallel implementations of Finite Element\nand Boundary Integral methods over curvilinear tetrahedral geometries,\nincluding symbolic polynomial mappings and operations, recursive integration,\nsparse and dense grid communication, parallel timing and memory footprint\ndiagnostics utilities. It is written in templated C++ using MPI for\nparallelization and ParMETIS for grid partitioning, and is provided as a module\nfor the DUNE interface. The dune-curvilineargrid grid manager is continuously\ndeveloped and improved, and so is this documentation. For the most recent\nversion of the documentation, as well as the source code, please refer to the\nprovided repositories and our website.\n", " The purpose of this communication is to suggest that the mechanism for\ngenerating the geomagnetic field and the energy source for powering it are one\nand the same, a nuclear georeactor at the center of the Earth. Toward this end,\nI: i) Present evidence that the nuclear georeactor fission-product sub-shell is\nfluid; ii)Suggest that the geomagnetic field is generated within the georeactor\nsub-shell, rather than within Earth's iron-alloy fluid core; iii) Describe why\nconvection appears more feasible within the georeactor sub-shell than within\nthe iron-alloy core; iv) Disclose additional relative physical advantages for\ngeoreactor sub-shell dynamo operation; and, v) Outline briefly the research\nthat should be conducted to advance the state of knowledge of\ngeoreactor-geomagnetic field generation. The concept of geomagnetic field\nproduction by the nuclear georeactor is presented specifically for the Earth.\nThe concepts and principles, however, are generally applicable to planetary\nmagnetic field production.\n", " We demonstrate that the classification of boosted, hadronically-decaying weak\ngauge bosons can be significantly improved over traditional cut-based and\nBDT-based methods using deep learning and the jet charge variable. We construct\nbinary taggers for $W^+$ vs. $W^-$ and $Z$ vs. $W$ discrimination, as well as\nan overall ternary classifier for $W^+$/$W^-$/$Z$ discrimination. Besides a\nsimple convolutional neural network (CNN), we also explore a composite of two\nCNNs, with different numbers of layers in the jet $p_{T}$ and jet charge\nchannels. We find that this novel structure boosts the performance particularly\nwhen considering the $Z$ boson as signal. The methods presented here can\nenhance the physics potential in SM measurements and searches for new physics\nthat are sensitive to the electric charge of weak gauge bosons.\n", " Perhaps the simplest first-principles approach to electronic structure is to\nfit the charge distribution of each orbital pair and use those fits wherever\nthey appear in the entire electron-electron (EE) interaction energy. The charge\ndistributions in quantum chemistry are typically represented as a sums over\nproducts of Gaussian orbital basis functions. If fitted, they are also\nrepresented as a sum over single-center Gaussian fitting basis functions. With\ntwo representations of the charge distributions, the proper definition of\nenergy is ambiguous. To remedy this, we require that the variation of the\nenergy with respect to a product of orbitals generates a fitted potential. This\nmakes the quantum-mechanical energy robust, i.e. corrected to first order for\nthe error made using an incomplete fitting basis. The coupled orbital and\nfitting equations are then the result of making the energy stationary with\nrespect to two independent sets of variables. We define the potentials and\nunique energies for methods based on the Hartree Fock model and variationally\nfit the full EE interaction in DFT. We compare implementations of variational\nfitting in DFT at six different levels for three different functionals. Our\ncalculations are performed on transition metal atoms, for which first-order\nCoulomb errors, due to an incomplete fitting basis sets, are significant.\nVariational first-order exchange and correlation errors have similar magnitude\nin all cases. Robust energy differences are much smaller, particularly in the\nlocal density approximation.\n", " We consider dynamics of excitons in branched conducting polymers. An\neffective model based on the use of quantum graph concept is applied for\ncomputing of exciton migration along the branched polymer chain Condition for\nthe regime, when the transmission of exciton through the branching point is\nreflectionless is revealed.\n", " Fault tolerance is increasingly important for unmanned autonomous vehicles.\nFor example, in a multi robot system the agents need the ability to effectively\ndetect and tolerate internal failures in order to continue performing their\ntasks without the need for immediate human intervention. The system must react\nto unplanned events in order to optimize the task allocation between the\nrobots. In a broad sense, the resilience of a system can be defined as the\nability to maintain or recover a stable state when subject to disturbance and\nit is related to the concept of robustness in industrial systems. In this\npaper, we study the resilience in a synchronized multi-robot system stated as\nfollows:Consider a team of $n$ (ground or aerial) robots each moving along\npredetermined periodic closed trajectories. Each of the agents needs to\ncommunicate informationabout its operation to other agents, but the\ncommunication links have a limited range. Hence, when two agents are within\ncommunication range, a communication link is established, and information is\nexchanged. Thus, two neighbors are synchronized if they visit the communication\nlink at the same time and a multi-robot system is called synchronized if each\npair of neighbors is synchronized. If a set of robots left the system, then\nsome trajectories has no robots. In these cases, when an alive robot detects no\nneighboring robot then it pass to this neighboring trajectory to assume the\nunattended task. In this framework, a fault-tolerance measure is introduced:\nthe resilience of the system is the largest number of robots that can fail\nwhile executing the global task. Interesting combinatorial properties of the\nresilience are showed that allow to know its value for some usual scenarios.\n", " The protein p53 has a well established role in protecting genomic integrity\nin human cells. When DNA is damaged p53 induces the cell cycle arrest to\nprevent the transmission of the damage to cell progeny, triggers the production\nof proteins for DNA repair and ultimately calls for apoptosis. In particular,\nthe p53-Mdm2 feedback loop seems to be the key circuit in this response of\ncells to damage. For many years, based on measurements over populations of\ncells it was believed that the p53-Mdm2 feedback loop was the responsible for\nthe existence of damped oscillations in the levels of p53 and Mdm2 after DNA\ndamage. However, recent measurements in individual human cells have shown that\np53 and its regulator Mdm2 develop sustained oscillations over long periods of\ntime even in the absence of stress. These results have attracted a lot of\ninterest, first because they open a new experimental framework to study the p53\nand its interactions and second because they challenge years of mathematical\nmodels with new and accurate data on single cells. Inspired by these\nexperiments standard models of the p53-Mdm2 circuit were modified introducing\nad-hoc some biologically motivated noise that becomes responsible for the\nstability of the oscillations. Here, we follow an alternative approach\nproposing that the noise that stabilizes the fluctuations is the intrinsic\nnoise due to the finite nature of the populations of p53 and Mdm2 in a single\ncell.\n", " Electron transport in realistic physical and chemical systems often involves\nthe non-trivial exchange of energy with a large environment, requiring the\ndefinition and treatment of open quantum systems. Because the time evolution of\nan open quantum system employs a non-unitary operator, the simulation of open\nquantum systems presents a challenge for universal quantum computers\nconstructed from only unitary operators or gates. Here we present a general\nalgorithm for implementing the action of any non-unitary operator on an\narbitrary state on a quantum device. We show that any quantum operator can be\nexactly decomposed as a linear combination of at most four unitary operators.\nWe demonstrate this method on a two-level system in both zero and finite\ntemperature amplitude damping channels. The results are in agreement with\nclassical calculations, showing promise in simulating non-unitary operations on\nintermediate-term and future quantum devices.\n", " A parallel algorithm has perfect strong scaling if its running time on P\nprocessors is linear in 1/P, including all communication costs.\nDistributed-memory parallel algorithms for matrix multiplication with perfect\nstrong scaling have only recently been found. One is based on classical matrix\nmultiplication (Solomonik and Demmel, 2011), and one is based on Strassen's\nfast matrix multiplication (Ballard, Demmel, Holtz, Lipshitz, and Schwartz,\n2012). Both algorithms scale perfectly, but only up to some number of\nprocessors where the inter-processor communication no longer scales.\n We obtain a memory-independent communication cost lower bound on classical\nand Strassen-based distributed-memory matrix multiplication algorithms. These\nbounds imply that no classical or Strassen-based parallel matrix multiplication\nalgorithm can strongly scale perfectly beyond the ranges already attained by\nthe two parallel algorithms mentioned above. The memory-independent bounds and\nthe strong scaling bounds generalize to other algorithms.\n", " Movable antenna (MA) is a promising technology to improve wireless\ncommunication performance by varying the antenna position in a given finite\narea at the transceivers to create more favorable channel conditions. In this\npaper, we investigate the MA-enhanced multiple-access channel (MAC) for the\nuplink transmission from multiple users each equipped with a single MA to a\nbase station (BS) with a fixed-position antenna (FPA) array. A field-response\nbased channel model is used to characterize the multi-path channel between the\nantenna array of the BS and each user's MA with a flexible position. To\nevaluate the MAC performance gain provided by MAs, we formulate an optimization\nproblem for minimizing the total transmit power of users, subject to a\nminimum-achievable-rate requirement for each user, where the positions of MAs\nand the transmit powers of users, as well as the receive combining matrix of\nthe BS are jointly optimized. To solve this non-convex optimization problem\ninvolving intricately coupled variables, we develop two algorithms based on\nzero-forcing (ZF) and minimum mean square error (MMSE) combining methods,\nrespectively. Specifically, for each algorithm, the combining matrix of the BS\nand the total transmit power of users are expressed as a function of the MAs'\nposition vectors, which are then optimized by using the proposed\nmulti-directional descent (MDD) framework. It is shown that the proposed\nZF-based and MMSE-based MDD algorithms can converge to high-quality suboptimal\nsolutions with low computational complexities. Simulation results demonstrate\nthat the proposed solutions for MA-enhanced multiple access systems can\nsignificantly decrease the total transmit power of users as compared to\nconventional FPA systems employing antenna selection under both perfect and\nimperfect field-response information.\n", " We consider wave propagation and scattering governed by one dimensional\nSchrodinger operators with truncated periodic potentials. The propagation of\nwave packets with narrow frequency supports is studied. The goal is to describe\npotentials for which the group velocity (for periodic problem) is small and the\ntransmission coefficient for the truncated potential is not too small, i.e. to\nfind media where a slowing down of the wave packets coexists with a\ntransparency.\n", " Mass spectrometry (MS) based omics data analysis require significant time and\nresources. To date, few parallel algorithms have been proposed for deducing\npeptides from mass spectrometry-based data. However, these parallel algorithms\nwere designed, and developed when the amount of data that needed to be\nprocessed was smaller in scale. In this paper, we prove that the communication\nbound that is reached by the \\emph{existing} parallel algorithms is\n$\\Omega(mn+2r\\frac{q}{p})$, where $m$ and $n$ are the dimensions of the\ntheoretical database matrix, $q$ and $r$ are dimensions of spectra, and $p$ is\nthe number of processors. We further prove that communication-optimal strategy\nwith fast-memory $\\sqrt{M} = mn + \\frac{2qr}{p}$ can achieve\n$\\Omega({\\frac{2mnq}{p}})$ but is not achieved by any existing parallel\nproteomics algorithms till date. To validate our claim, we performed a\nmeta-analysis of published parallel algorithms, and their performance results.\nWe show that sub-optimal speedups with increasing number of processors is a\ndirect consequence of not achieving the communication lower-bounds. We further\nvalidate our claim by performing experiments which demonstrate the\ncommunication bounds that are proved in this paper. Consequently, we assert\nthat next-generation of \\emph{provable}, and demonstrated superior parallel\nalgorithms are urgently needed for MS based large systems-biology studies\nespecially for meta-proteomics, proteogenomic, microbiome, and proteomics for\nnon-model organisms. Our hope is that this paper will excite the parallel\ncomputing community to further investigate parallel algorithms for highly\ninfluential MS based omics problems.\n", " Here I design an experimental way of a quantum communication by quantum CNOT\ngates and single qubit gates without the help of classical communication.\n", " We prove new exponents for the energy version of the Erd\\H{o}s-Szemer\\'edi\nsum-product conjecture, raised by Balog and Wooley. They match the previously\nestablished milestone values for the standard formulation of the question, both\nfor general fields and the special case of real or complex numbers, and appear\nto be the best ones attainable within the currently available technology.\nFurther results are obtained about multiplicative energies of additive shifts\nand a strengthened energy version of the \"few sums, many products\" inequality\nof Elekes and Ruzsa. The latter inequality enables us to obtain a minor\nimprovement of the state-of the art sum-product exponent over the reals due to\nKonyagin and the second author, up to $\\frac{4}{3}+\\frac{1}{1509}$. An\napplication of energy estimates to an instance of arithmetic growth in prime\nresidue fields is presented.\n", " Retrieving unlabeled videos by textual queries, known as Ad-hoc Video Search\n(AVS), is a core theme in multimedia data management and retrieval. The success\nof AVS counts on cross-modal representation learning that encodes both query\nsentences and videos into common spaces for semantic similarity computation.\nInspired by the initial success of previously few works in combining multiple\nsentence encoders, this paper takes a step forward by developing a new and\ngeneral method for effectively exploiting diverse sentence encoders. The\nnovelty of the proposed method, which we term Sentence Encoder Assembly (SEA),\nis two-fold. First, different from prior art that use only a single common\nspace, SEA supports text-video matching in multiple encoder-specific common\nspaces. Such a property prevents the matching from being dominated by a\nspecific encoder that produces an encoding vector much longer than other\nencoders. Second, in order to explore complementarities among the individual\ncommon spaces, we propose multi-space multi-loss learning. As extensive\nexperiments on four benchmarks (MSR-VTT, TRECVID AVS 2016-2019, TGIF and MSVD)\nshow, SEA surpasses the state-of-the-art. In addition, SEA is extremely ease to\nimplement. All this makes SEA an appealing solution for AVS and promising for\ncontinuously advancing the task by harvesting new sentence encoders.\n", " One of the first widespread uses of multi-user multiple-input multiple-output\n(MU-MIMO) is in 5G networks, where each base station has an advanced antenna\nsystem (AAS) that is connected to the baseband unit (BBU) with a\ncapacity-constrained fronthaul. In the AAS configuration, multiple passive\nantenna elements and radio units are integrated into a single box. This paper\nconsiders precoded downlink transmission over a single-cell MU-MIMO system. We\nstudy optimized linear precoding for AAS with a limited-capacity fronthaul,\nwhich requires the precoding matrix to be quantized. We propose a new precoding\ndesign that is aware of the fronthaul quantization and minimizes the\nmean-squared error at the receiver side. We compute the precoding matrix using\na sphere decoding (SD) approach. We also propose a heuristic low-complexity\napproach to quantized precoding. This heuristic is computationally efficient\nenough for massive MIMO systems. The numerical results show that our proposed\nprecoding significantly outperforms quantization-unaware precoding and other\nprevious approaches in terms of the sum rate. The performance loss for our\nheuristic method compared to quantization-aware precoding is insignificant\nconsidering the complexity reduction, which makes the heuristic method feasible\nfor real-time applications. We consider both perfect and imperfect channel\nstate information.\n", " We study the four channels associated with neutrino-deuteron breakup\nreactions at next-to-next to leading order in effective field theory. We find\nthat the total cross-section is indeed converging for neutrino energies up to\n20 MeV, and thus our calculations can provide constraints on theoretical\nuncertainties for the Sudbury Neutrino Observatory. We stress the importance of\na direct experimental measurement to high precision in at least one channel, in\norder to fix an axial two-body counterterm.\n", " We address, through the automated reasoning tools in GeoGebra Discovery, a\nproblem from a regional phase of the Austrian Mathematics Olympiad 2023. Trying\nto solve this problem gives rise to four different kind of feedback: the almost\ninstantaneous, automated solution of the proposed problem; the measure of its\ncomplexity, according to some recent proposals; the automated discovery of a\ngeneralization of the given assertion, showing that the same statement is true\nover more general polygons than those mentioned in the problem; and the\ndifficulties associated to the analysis of the surprising and involved high\nnumber of degenerate cases that appear when using the LocusEquation command in\nthis problem. In our communication we will describe and reflect on these\ndiverse issues, enhancing its exemplar role for showing some of the advantages,\nproblems, and current fields of development of GeoGebra Discovery.\n", " This work investigates spectrum and root functions (that is, eigen- and\nassociated functions) of a Sturm-Liouville problem involving an abstract linear\noperator (nonselfadjoint in general) in the equation together with\nsupplementary transmission conditions at the one interior singular point. So,\nthe problem under consideration is not pure differential pro lem. At first we\nestablish isomorphism and coerciveness with respect to the spectral parameter\nfor the corresponding nonhomogeneous problem. Then by suggesting an own our\nmethod we prove that the spectrum of the considered problem is discrete and\nderive an asymptotic approximation formulas for the eigenvalues. We must note\nthat Asymptotics of the eigenvalues of such type problems is investigated at\nfirst in literature in the present work and the obtained results are new even\nin the continuous case (i.e. without transmission conditions). Finally it is\nshown that the system of root functions form an Abel basis of the corresponding\nHilbert space.\n", " The effect of signal integration through cis-regulatory modules (CRMs) on\nsynchronization and clustering of populations of two-component genetic\noscillators coupled by quorum sensing is in detail investigated. We find that\nthe CRMs play an important role in achieving synchronization and clustering.\nFor this, we investigate 6 possible cis-regulatory input functions (CRIFs) with\nAND, OR, ANDN, ORN, XOR, and EQU types of responses in two possible kinds of\ncell-to-cell communications: activator-regulated communication (i.e., the\nautoinducer regulates the activator) and repressor-regulated communication\n(i.e., the autoinducer regulates the repressor). Both theoretical analysis and\nnumerical simulation show that different CRMs drive fundamentally different\ncellular patterns, such as complete synchronization, various cluster-balanced\nstates and several cluster-nonbalanced states.\n", " Many quantum algorithms, including recently proposed hybrid classical/quantum\nalgorithms, make use of restricted tomography of the quantum state that\nmeasures the reduced density matrices, or marginals, of the full state. The\nmost straightforward approach to this algorithmic step estimates each component\nof the marginal independently without making use of the algebraic and geometric\nstructure of the marginals. Within the field of quantum chemistry, this\nstructure is termed the fermionic $n$-representability conditions, and is\nsupported by a vast amount of literature on both theoretical and practical\nresults related to their approximations. In this work, we introduce these\nconditions in the language of quantum computation, and utilize them to develop\nseveral techniques to accelerate and improve practical applications for quantum\nchemistry on quantum computers. We show that one can use fermionic\n$n$-representability conditions to reduce the total number of measurements\nrequired by more than an order of magnitude for medium sized systems in\nchemistry. We also demonstrate an efficient restoration of the physicality of\nenergy curves for the dilation of a four qubit diatomic hydrogen system in the\npresence of three distinct one qubit error channels, providing evidence these\ntechniques are useful for pre-fault tolerant quantum chemistry experiments.\n", " A simple model for the transmission of pairs of electrons through a weak\nelectric link in the form of a nanowire made of a material with strong electron\nspin-orbit interaction (SOI) is presented, with emphasis on the effects of\nCoulomb interactions and the Pauli exclusion principle. The constraints due to\nthe Pauli principle are shown to \"quench\" the coherent SOI-induced precession\nof the spins when the spatial wave packets of the two electrons overlap\nsignificantly. The quenching, which results from the projection of the pair's\nspin states onto spin-up and spin-down states on the link, breaks up the\ncoherent propagation in the link into a sequence of coherent hops that add\nincoherently. Applying the model to the transmission of Cooper pairs between\ntwo superconductors, we find that in spite of Pauli quenching, the Josephson\ncurrent oscillates with the strength of the SOI, and may even change its sign.\nConditions for an experimental detection of these features are discussed.\n", " Here, we propose a novel approach to experimentally and theoretically study\nthe properties of QCD matter under new extreme conditions, namely having an\ninitial temperature over 300~MeV and baryonic charge density over three times\nthe values of the normal nuclear density. According to contemporary theoretical\nknowledge, such conditions were not accessible during the early Universe\nevolution and are not accessible now in the known astrophysical phenomena. To\nachieve these new extreme conditions, we proposed performing high-luminosity\nexperiments at LHC or other colliders by means of scattering the two colliding\nbeams at the nuclei of a solid target that is fixed at their interaction\nregion. Under plausible assumptions, we estimate the reaction rate for the\np+C+p and Pb+Pb+Pb reactions and discuss the energy deposition into the target\nand possible types of fixed targets for such reactions. To simulate the triple\nnuclear collisions, we employed the well-known UrQMD 3.4 model for the beam\ncenter-of-mass collision energies 2.76 TeV. As a result of our modeling, we\nfound that, in the most central and simultaneous triple nuclear collisions, the\ninitial baryonic charge density is approximately three times higher than the\none achieved in the ordinary binary nuclear collisions at this energy.\n" ]