id
large_string | submitter
large_string | authors
large_string | title
large_string | comments
large_string | journal-ref
large_string | doi
large_string | report-no
large_string | categories
large_string | license
large_string | abstract
large_string | versions
large list | update_date
timestamp[ms] | authors_parsed
large list | prompt
large_string |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
1411.3811 | Todor Ivanov | Todor Ivanov, Roberto V. Zicari, Sead Izberovic, Karsten Tolle | Performance Evaluation of Virtualized Hadoop Clusters | null | null | null | Technical Report No. 2014-1 | cs.DC cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this report we investigate the performance of Hadoop clusters, deployed
with separated storage and compute layers, on top of a hypervisor managing a
single physical host. We have analyzed and evaluated the different Hadoop
cluster configurations by running CPU bound and I/O bound workloads. The report
is structured as follows: Section 2 provides a brief description of the
technologies involved in our study. An overview of the experimental platform,
setup test and configurations are presented in Section 3. Our benchmark
methodology is defined in Section 4. The performed experiments together with
the evaluation of the results are presented in Section 5. Finally, Section 6
concludes with lessons learned.
| [
{
"version": "v1",
"created": "Fri, 14 Nov 2014 07:14:05 GMT"
}
] | 2014-11-17T00:00:00 | [
[
"Ivanov",
"Todor",
""
],
[
"Zicari",
"Roberto V.",
""
],
[
"Izberovic",
"Sead",
""
],
[
"Tolle",
"Karsten",
""
]
] | TITLE: Performance Evaluation of Virtualized Hadoop Clusters
ABSTRACT: In this report we investigate the performance of Hadoop clusters, deployed
with separated storage and compute layers, on top of a hypervisor managing a
single physical host. We have analyzed and evaluated the different Hadoop
cluster configurations by running CPU bound and I/O bound workloads. The report
is structured as follows: Section 2 provides a brief description of the
technologies involved in our study. An overview of the experimental platform,
setup test and configurations are presented in Section 3. Our benchmark
methodology is defined in Section 4. The performed experiments together with
the evaluation of the results are presented in Section 5. Finally, Section 6
concludes with lessons learned.
|
1411.4006 | Zhongwen Xu | Zhongwen Xu, Yi Yang and Alexander G. Hauptmann | A Discriminative CNN Video Representation for Event Detection | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we propose a discriminative video representation for event
detection over a large scale video dataset when only limited hardware resources
are available. The focus of this paper is to effectively leverage deep
Convolutional Neural Networks (CNNs) to advance event detection, where only
frame level static descriptors can be extracted by the existing CNN toolkit.
This paper makes two contributions to the inference of CNN video
representation. First, while average pooling and max pooling have long been the
standard approaches to aggregating frame level static features, we show that
performance can be significantly improved by taking advantage of an appropriate
encoding method. Second, we propose using a set of latent concept descriptors
as the frame descriptor, which enriches visual information while keeping it
computationally affordable. The integration of the two contributions results in
a new state-of-the-art performance in event detection over the largest video
datasets. Compared to improved Dense Trajectories, which has been recognized as
the best video representation for event detection, our new representation
improves the Mean Average Precision (mAP) from 27.6% to 36.8% for the TRECVID
MEDTest 14 dataset and from 34.0% to 44.6% for the TRECVID MEDTest 13 dataset.
This work is the core part of the winning solution of our CMU-Informedia team
in TRECVID MED 2014 competition.
| [
{
"version": "v1",
"created": "Fri, 14 Nov 2014 18:37:31 GMT"
}
] | 2014-11-17T00:00:00 | [
[
"Xu",
"Zhongwen",
""
],
[
"Yang",
"Yi",
""
],
[
"Hauptmann",
"Alexander G.",
""
]
] | TITLE: A Discriminative CNN Video Representation for Event Detection
ABSTRACT: In this paper, we propose a discriminative video representation for event
detection over a large scale video dataset when only limited hardware resources
are available. The focus of this paper is to effectively leverage deep
Convolutional Neural Networks (CNNs) to advance event detection, where only
frame level static descriptors can be extracted by the existing CNN toolkit.
This paper makes two contributions to the inference of CNN video
representation. First, while average pooling and max pooling have long been the
standard approaches to aggregating frame level static features, we show that
performance can be significantly improved by taking advantage of an appropriate
encoding method. Second, we propose using a set of latent concept descriptors
as the frame descriptor, which enriches visual information while keeping it
computationally affordable. The integration of the two contributions results in
a new state-of-the-art performance in event detection over the largest video
datasets. Compared to improved Dense Trajectories, which has been recognized as
the best video representation for event detection, our new representation
improves the Mean Average Precision (mAP) from 27.6% to 36.8% for the TRECVID
MEDTest 14 dataset and from 34.0% to 44.6% for the TRECVID MEDTest 13 dataset.
This work is the core part of the winning solution of our CMU-Informedia team
in TRECVID MED 2014 competition.
|
cs/9508101 | null | Q. Zhao, T. Nishida | Using Qualitative Hypotheses to Identify Inaccurate Data | See http://www.jair.org/ for any accompanying files | Journal of Artificial Intelligence Research, Vol 3, (1995),
119-145 | null | null | cs.AI | null | Identifying inaccurate data has long been regarded as a significant and
difficult problem in AI. In this paper, we present a new method for identifying
inaccurate data on the basis of qualitative correlations among related data.
First, we introduce the definitions of related data and qualitative
correlations among related data. Then we put forward a new concept called
support coefficient function (SCF). SCF can be used to extract, represent, and
calculate qualitative correlations among related data within a dataset. We
propose an approach to determining dynamic shift intervals of inaccurate data,
and an approach to calculating possibility of identifying inaccurate data,
respectively. Both of the approaches are based on SCF. Finally we present an
algorithm for identifying inaccurate data by using qualitative correlations
among related data as confirmatory or disconfirmatory evidence. We have
developed a practical system for interpreting infrared spectra by applying the
method, and have fully tested the system against several hundred real spectra.
The experimental results show that the method is significantly better than the
conventional methods used in many similar systems.
| [
{
"version": "v1",
"created": "Tue, 1 Aug 1995 00:00:00 GMT"
}
] | 2014-11-17T00:00:00 | [
[
"Zhao",
"Q.",
""
],
[
"Nishida",
"T.",
""
]
] | TITLE: Using Qualitative Hypotheses to Identify Inaccurate Data
ABSTRACT: Identifying inaccurate data has long been regarded as a significant and
difficult problem in AI. In this paper, we present a new method for identifying
inaccurate data on the basis of qualitative correlations among related data.
First, we introduce the definitions of related data and qualitative
correlations among related data. Then we put forward a new concept called
support coefficient function (SCF). SCF can be used to extract, represent, and
calculate qualitative correlations among related data within a dataset. We
propose an approach to determining dynamic shift intervals of inaccurate data,
and an approach to calculating possibility of identifying inaccurate data,
respectively. Both of the approaches are based on SCF. Finally we present an
algorithm for identifying inaccurate data by using qualitative correlations
among related data as confirmatory or disconfirmatory evidence. We have
developed a practical system for interpreting infrared spectra by applying the
method, and have fully tested the system against several hundred real spectra.
The experimental results show that the method is significantly better than the
conventional methods used in many similar systems.
|
1411.3374 | Manas Joglekar | Manas Joglekar, Hector Garcia-Molina, Aditya Parameswaran, Christopher
Re | Exploiting Correlations for Expensive Predicate Evaluation | null | null | null | null | cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | User Defined Function(UDFs) are used increasingly to augment query languages
with extra, application dependent functionality. Selection queries involving
UDF predicates tend to be expensive, either in terms of monetary cost or
latency. In this paper, we study ways to efficiently evaluate selection queries
with UDF predicates. We provide a family of techniques for processing queries
at low cost while satisfying user-specified precision and recall constraints.
Our techniques are applicable to a variety of scenarios including when
selection probabilities of tuples are available beforehand, when this
information is available but noisy, or when no such prior information is
available. We also generalize our techniques to more complex queries. Finally,
we test our techniques on real datasets, and show that they achieve significant
savings in cost of up to $80\%$, while incurring only a small reduction in
accuracy.
| [
{
"version": "v1",
"created": "Wed, 12 Nov 2014 22:00:08 GMT"
}
] | 2014-11-14T00:00:00 | [
[
"Joglekar",
"Manas",
""
],
[
"Garcia-Molina",
"Hector",
""
],
[
"Parameswaran",
"Aditya",
""
],
[
"Re",
"Christopher",
""
]
] | TITLE: Exploiting Correlations for Expensive Predicate Evaluation
ABSTRACT: User Defined Function(UDFs) are used increasingly to augment query languages
with extra, application dependent functionality. Selection queries involving
UDF predicates tend to be expensive, either in terms of monetary cost or
latency. In this paper, we study ways to efficiently evaluate selection queries
with UDF predicates. We provide a family of techniques for processing queries
at low cost while satisfying user-specified precision and recall constraints.
Our techniques are applicable to a variety of scenarios including when
selection probabilities of tuples are available beforehand, when this
information is available but noisy, or when no such prior information is
available. We also generalize our techniques to more complex queries. Finally,
we test our techniques on real datasets, and show that they achieve significant
savings in cost of up to $80\%$, while incurring only a small reduction in
accuracy.
|
1411.3377 | Manas Joglekar | Manas Joglekar, Hector Garcia-Molina, Aditya Parameswaran | Comprehensive and Reliable Crowd Assessment Algorithms | ICDE 2015 | null | null | null | cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Evaluating workers is a critical aspect of any crowdsourcing system. In this
paper, we devise techniques for evaluating workers by finding confidence
intervals on their error rates. Unlike prior work, we focus on
"conciseness"---that is, giving as tight a confidence interval as possible.
Conciseness is of utmost importance because it allows us to be sure that we
have the best guarantee possible on worker error rate. Also unlike prior work,
we provide techniques that work under very general scenarios, such as when not
all workers have attempted every task (a fairly common scenario in practice),
when tasks have non-boolean responses, and when workers have different biases
for positive and negative tasks. We demonstrate conciseness as well as accuracy
of our confidence intervals by testing them on a variety of conditions and
multiple real-world datasets.
| [
{
"version": "v1",
"created": "Wed, 12 Nov 2014 22:09:17 GMT"
}
] | 2014-11-14T00:00:00 | [
[
"Joglekar",
"Manas",
""
],
[
"Garcia-Molina",
"Hector",
""
],
[
"Parameswaran",
"Aditya",
""
]
] | TITLE: Comprehensive and Reliable Crowd Assessment Algorithms
ABSTRACT: Evaluating workers is a critical aspect of any crowdsourcing system. In this
paper, we devise techniques for evaluating workers by finding confidence
intervals on their error rates. Unlike prior work, we focus on
"conciseness"---that is, giving as tight a confidence interval as possible.
Conciseness is of utmost importance because it allows us to be sure that we
have the best guarantee possible on worker error rate. Also unlike prior work,
we provide techniques that work under very general scenarios, such as when not
all workers have attempted every task (a fairly common scenario in practice),
when tasks have non-boolean responses, and when workers have different biases
for positive and negative tasks. We demonstrate conciseness as well as accuracy
of our confidence intervals by testing them on a variety of conditions and
multiple real-world datasets.
|
1411.3409 | Paul Mineiro | Paul Mineiro, Nikos Karampatziakis | A Randomized Algorithm for CCA | null | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present RandomizedCCA, a randomized algorithm for computing canonical
analysis, suitable for large datasets stored either out of core or on a
distributed file system. Accurate results can be obtained in as few as two data
passes, which is relevant for distributed processing frameworks in which
iteration is expensive (e.g., Hadoop). The strategy also provides an excellent
initializer for standard iterative solutions.
| [
{
"version": "v1",
"created": "Thu, 13 Nov 2014 00:51:19 GMT"
}
] | 2014-11-14T00:00:00 | [
[
"Mineiro",
"Paul",
""
],
[
"Karampatziakis",
"Nikos",
""
]
] | TITLE: A Randomized Algorithm for CCA
ABSTRACT: We present RandomizedCCA, a randomized algorithm for computing canonical
analysis, suitable for large datasets stored either out of core or on a
distributed file system. Accurate results can be obtained in as few as two data
passes, which is relevant for distributed processing frameworks in which
iteration is expensive (e.g., Hadoop). The strategy also provides an excellent
initializer for standard iterative solutions.
|
1411.3410 | Kwangchol Jang | Kwangchol Jang, Sokmin Han, Insong Kim | Person Re-identification Based on Color Histogram and Spatial
Configuration of Dominant Color Regions | 12 pages, 6 figures | null | null | null | cs.CV | http://creativecommons.org/licenses/publicdomain/ | There is a requirement to determine whether a given person of interest has
already been observed over a network of cameras in video surveillance systems.
A human appearance obtained in one camera is usually different from the ones
obtained in another camera due to difference in illumination, pose and
viewpoint, camera parameters. Being related to appearance-based approaches for
person re-identification, we propose a novel method based on the dominant color
histogram and spatial configuration of dominant color regions on human body
parts. Dominant color histogram and spatial configuration of the dominant color
regions based on dominant color descriptor(DCD) can be considered to be robust
to illumination and pose, viewpoint changes. The proposed method is evaluated
using benchmark video datasets. Experimental results using the cumulative
matching characteristic(CMC) curve demonstrate the effectiveness of our
approach for person re-identification.
| [
{
"version": "v1",
"created": "Thu, 13 Nov 2014 00:55:48 GMT"
}
] | 2014-11-14T00:00:00 | [
[
"Jang",
"Kwangchol",
""
],
[
"Han",
"Sokmin",
""
],
[
"Kim",
"Insong",
""
]
] | TITLE: Person Re-identification Based on Color Histogram and Spatial
Configuration of Dominant Color Regions
ABSTRACT: There is a requirement to determine whether a given person of interest has
already been observed over a network of cameras in video surveillance systems.
A human appearance obtained in one camera is usually different from the ones
obtained in another camera due to difference in illumination, pose and
viewpoint, camera parameters. Being related to appearance-based approaches for
person re-identification, we propose a novel method based on the dominant color
histogram and spatial configuration of dominant color regions on human body
parts. Dominant color histogram and spatial configuration of the dominant color
regions based on dominant color descriptor(DCD) can be considered to be robust
to illumination and pose, viewpoint changes. The proposed method is evaluated
using benchmark video datasets. Experimental results using the cumulative
matching characteristic(CMC) curve demonstrate the effectiveness of our
approach for person re-identification.
|
1406.2199 | Karen Simonyan | Karen Simonyan, Andrew Zisserman | Two-Stream Convolutional Networks for Action Recognition in Videos | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate architectures of discriminatively trained deep Convolutional
Networks (ConvNets) for action recognition in video. The challenge is to
capture the complementary information on appearance from still frames and
motion between frames. We also aim to generalise the best performing
hand-crafted features within a data-driven learning framework.
Our contribution is three-fold. First, we propose a two-stream ConvNet
architecture which incorporates spatial and temporal networks. Second, we
demonstrate that a ConvNet trained on multi-frame dense optical flow is able to
achieve very good performance in spite of limited training data. Finally, we
show that multi-task learning, applied to two different action classification
datasets, can be used to increase the amount of training data and improve the
performance on both.
Our architecture is trained and evaluated on the standard video actions
benchmarks of UCF-101 and HMDB-51, where it is competitive with the state of
the art. It also exceeds by a large margin previous attempts to use deep nets
for video classification.
| [
{
"version": "v1",
"created": "Mon, 9 Jun 2014 14:44:14 GMT"
},
{
"version": "v2",
"created": "Wed, 12 Nov 2014 20:48:33 GMT"
}
] | 2014-11-13T00:00:00 | [
[
"Simonyan",
"Karen",
""
],
[
"Zisserman",
"Andrew",
""
]
] | TITLE: Two-Stream Convolutional Networks for Action Recognition in Videos
ABSTRACT: We investigate architectures of discriminatively trained deep Convolutional
Networks (ConvNets) for action recognition in video. The challenge is to
capture the complementary information on appearance from still frames and
motion between frames. We also aim to generalise the best performing
hand-crafted features within a data-driven learning framework.
Our contribution is three-fold. First, we propose a two-stream ConvNet
architecture which incorporates spatial and temporal networks. Second, we
demonstrate that a ConvNet trained on multi-frame dense optical flow is able to
achieve very good performance in spite of limited training data. Finally, we
show that multi-task learning, applied to two different action classification
datasets, can be used to increase the amount of training data and improve the
performance on both.
Our architecture is trained and evaluated on the standard video actions
benchmarks of UCF-101 and HMDB-51, where it is competitive with the state of
the art. It also exceeds by a large margin previous attempts to use deep nets
for video classification.
|
1411.3041 | Ramakrishna Vedantam | Ramakrishna Vedantam, C. Lawrence Zitnick, and Devi Parikh | Collecting Image Description Datasets using Crowdsourcing | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We describe our two new datasets with images described by humans. Both the
datasets were collected using Amazon Mechanical Turk, a crowdsourcing platform.
The two datasets contain significantly more descriptions per image than other
existing datasets. One is based on a popular image description dataset called
the UIUC Pascal Sentence Dataset, whereas the other is based on the Abstract
Scenes dataset con- taining images made from clipart objects. In this paper we
describe our interfaces, analyze some properties of and show example
descriptions from our two datasets.
| [
{
"version": "v1",
"created": "Wed, 12 Nov 2014 01:34:46 GMT"
}
] | 2014-11-13T00:00:00 | [
[
"Vedantam",
"Ramakrishna",
""
],
[
"Zitnick",
"C. Lawrence",
""
],
[
"Parikh",
"Devi",
""
]
] | TITLE: Collecting Image Description Datasets using Crowdsourcing
ABSTRACT: We describe our two new datasets with images described by humans. Both the
datasets were collected using Amazon Mechanical Turk, a crowdsourcing platform.
The two datasets contain significantly more descriptions per image than other
existing datasets. One is based on a popular image description dataset called
the UIUC Pascal Sentence Dataset, whereas the other is based on the Abstract
Scenes dataset con- taining images made from clipart objects. In this paper we
describe our interfaces, analyze some properties of and show example
descriptions from our two datasets.
|
1411.3201 | Swetha P T Srinivasan | Swetha P.T. Srinivasan, Umesh Bellur | Novel Power and Completion Time Models for Virtualized Environments | null | null | null | null | cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Power consumption costs takes upto half of operational expenses of
datacenters making power management a critical concern. Advances in processor
technology provide fine-grained control over operating frequency and voltage of
processors and this control can be used to tradeoff power for performance.
Although many power and performance models exist, they have a significant error
margin while predicting the performance of memory or file-intensive tasks and
HPC applications. Our investigations reveal that the prediction error is due in
part to the fact that they do not take frequency AND CPU variations account,
rather they just depend on the CPU by itself.
In this paper, we empirically derive power and completion time models using
linear regression with CPU utilization and operating frequency as parameters.
We validate our power model on several Intel and AMD processors by predicting
within 2-7% of measured power. We validate our completion time model using five
kernels of NASA Parallel Benchmark suite and five CPU, memory and
file-intensive benchmarks on four heterogeneous systems and predicting within
1-6% of observed performance. We then show how these models can be employed to
realize as much as 15% savings in power while delivering 44% better performance
for applications deployed in a virtualized environment.
| [
{
"version": "v1",
"created": "Wed, 12 Nov 2014 15:16:51 GMT"
}
] | 2014-11-13T00:00:00 | [
[
"Srinivasan",
"Swetha P. T.",
""
],
[
"Bellur",
"Umesh",
""
]
] | TITLE: Novel Power and Completion Time Models for Virtualized Environments
ABSTRACT: Power consumption costs takes upto half of operational expenses of
datacenters making power management a critical concern. Advances in processor
technology provide fine-grained control over operating frequency and voltage of
processors and this control can be used to tradeoff power for performance.
Although many power and performance models exist, they have a significant error
margin while predicting the performance of memory or file-intensive tasks and
HPC applications. Our investigations reveal that the prediction error is due in
part to the fact that they do not take frequency AND CPU variations account,
rather they just depend on the CPU by itself.
In this paper, we empirically derive power and completion time models using
linear regression with CPU utilization and operating frequency as parameters.
We validate our power model on several Intel and AMD processors by predicting
within 2-7% of measured power. We validate our completion time model using five
kernels of NASA Parallel Benchmark suite and five CPU, memory and
file-intensive benchmarks on four heterogeneous systems and predicting within
1-6% of observed performance. We then show how these models can be employed to
realize as much as 15% savings in power while delivering 44% better performance
for applications deployed in a virtualized environment.
|
1411.3212 | Francesco Lettich | Francesco Lettich, Salvatore Orlando, Claudio Silvestri and Christian
S. Jensen | Manycore processing of repeated range queries over massive moving
objects observations | null | null | null | null | cs.DB cs.DC cs.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The ability to timely process significant amounts of continuously updated
spatial data is mandatory for an increasing number of applications. Parallelism
enables such applications to face this data-intensive challenge and allows the
devised systems to feature low latency and high scalability. In this paper we
focus on a specific data-intensive problem, concerning the repeated processing
of huge amounts of range queries over massive sets of moving objects, where the
spatial extents of queries and objects are continuously modified over time. To
tackle this problem and significantly accelerate query processing we devise a
hybrid CPU/GPU pipeline that compresses data output and save query processing
work. The devised system relies on an ad-hoc spatial index leading to a problem
decomposition that results in a set of independent data-parallel tasks. The
index is based on a point-region quadtree space decomposition and allows to
tackle effectively a broad range of spatial object distributions, even those
very skewed. Also, to deal with the architectural peculiarities and limitations
of the GPUs, we adopt non-trivial GPU data structures that avoid the need of
locked memory accesses and favour coalesced memory accesses, thus enhancing the
overall memory throughput. To the best of our knowledge this is the first work
that exploits GPUs to efficiently solve repeated range queries over massive
sets of continuously moving objects, characterized by highly skewed spatial
distributions. In comparison with state-of-the-art CPU-based implementations,
our method highlights significant speedups in the order of 14x-20x, depending
on the datasets, even when considering very cheap GPUs.
| [
{
"version": "v1",
"created": "Wed, 12 Nov 2014 15:46:39 GMT"
}
] | 2014-11-13T00:00:00 | [
[
"Lettich",
"Francesco",
""
],
[
"Orlando",
"Salvatore",
""
],
[
"Silvestri",
"Claudio",
""
],
[
"Jensen",
"Christian S.",
""
]
] | TITLE: Manycore processing of repeated range queries over massive moving
objects observations
ABSTRACT: The ability to timely process significant amounts of continuously updated
spatial data is mandatory for an increasing number of applications. Parallelism
enables such applications to face this data-intensive challenge and allows the
devised systems to feature low latency and high scalability. In this paper we
focus on a specific data-intensive problem, concerning the repeated processing
of huge amounts of range queries over massive sets of moving objects, where the
spatial extents of queries and objects are continuously modified over time. To
tackle this problem and significantly accelerate query processing we devise a
hybrid CPU/GPU pipeline that compresses data output and save query processing
work. The devised system relies on an ad-hoc spatial index leading to a problem
decomposition that results in a set of independent data-parallel tasks. The
index is based on a point-region quadtree space decomposition and allows to
tackle effectively a broad range of spatial object distributions, even those
very skewed. Also, to deal with the architectural peculiarities and limitations
of the GPUs, we adopt non-trivial GPU data structures that avoid the need of
locked memory accesses and favour coalesced memory accesses, thus enhancing the
overall memory throughput. To the best of our knowledge this is the first work
that exploits GPUs to efficiently solve repeated range queries over massive
sets of continuously moving objects, characterized by highly skewed spatial
distributions. In comparison with state-of-the-art CPU-based implementations,
our method highlights significant speedups in the order of 14x-20x, depending
on the datasets, even when considering very cheap GPUs.
|
1407.4896 | Hang-Hyun Jo | Hang-Hyun Jo, Jari Saram\"aki, Robin I. M. Dunbar, and Kimmo Kaski | Spatial patterns of close relationships across the lifespan | 9 pages, 7 figures | Scientific Reports 4, 6988 (2014) | 10.1038/srep06988 | null | physics.soc-ph cs.SI physics.data-an | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The dynamics of close relationships is important for understanding the
migration patterns of individual life-courses. The bottom-up approach to this
subject by social scientists has been limited by sample size, while the more
recent top-down approach using large-scale datasets suffers from a lack of
detail about the human individuals. We incorporate the geographic and
demographic information of millions of mobile phone users with their
communication patterns to study the dynamics of close relationships and its
effect in their life-course migration. We demonstrate how the close age- and
sex-biased dyadic relationships are correlated with the geographic proximity of
the pair of individuals, e.g., young couples tend to live further from each
other than old couples. In addition, we find that emotionally closer pairs are
living geographically closer to each other. These findings imply that the
life-course framework is crucial for understanding the complex dynamics of
close relationships and their effect on the migration patterns of human
individuals.
| [
{
"version": "v1",
"created": "Fri, 18 Jul 2014 06:46:45 GMT"
},
{
"version": "v2",
"created": "Wed, 17 Sep 2014 06:12:54 GMT"
}
] | 2014-11-12T00:00:00 | [
[
"Jo",
"Hang-Hyun",
""
],
[
"Saramäki",
"Jari",
""
],
[
"Dunbar",
"Robin I. M.",
""
],
[
"Kaski",
"Kimmo",
""
]
] | TITLE: Spatial patterns of close relationships across the lifespan
ABSTRACT: The dynamics of close relationships is important for understanding the
migration patterns of individual life-courses. The bottom-up approach to this
subject by social scientists has been limited by sample size, while the more
recent top-down approach using large-scale datasets suffers from a lack of
detail about the human individuals. We incorporate the geographic and
demographic information of millions of mobile phone users with their
communication patterns to study the dynamics of close relationships and its
effect in their life-course migration. We demonstrate how the close age- and
sex-biased dyadic relationships are correlated with the geographic proximity of
the pair of individuals, e.g., young couples tend to live further from each
other than old couples. In addition, we find that emotionally closer pairs are
living geographically closer to each other. These findings imply that the
life-course framework is crucial for understanding the complex dynamics of
close relationships and their effect on the migration patterns of human
individuals.
|
1410.4393 | Christopher Herbon | Christopher Herbon | The HAWKwood Database | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a database consisting of wood pile images, which can be used as a
benchmark to evaluate the performance of wood pile detection and surveying
algorithms. We distinguish six database cate- gories which can be used for
different types of algorithms. Images of real and synthetic scenes are
provided, which consist of 7655 images divided into 354 data sets. Depending on
the category the data sets either include ground truth data or forestry
specific measurements with which algorithms may be compared.
| [
{
"version": "v1",
"created": "Thu, 16 Oct 2014 12:25:50 GMT"
},
{
"version": "v2",
"created": "Tue, 11 Nov 2014 07:07:30 GMT"
}
] | 2014-11-12T00:00:00 | [
[
"Herbon",
"Christopher",
""
]
] | TITLE: The HAWKwood Database
ABSTRACT: We present a database consisting of wood pile images, which can be used as a
benchmark to evaluate the performance of wood pile detection and surveying
algorithms. We distinguish six database cate- gories which can be used for
different types of algorithms. Images of real and synthetic scenes are
provided, which consist of 7655 images divided into 354 data sets. Depending on
the category the data sets either include ground truth data or forestry
specific measurements with which algorithms may be compared.
|
1411.2795 | Nitesh Kumar Chaudhary | Nitesh Kumar Chaudhary | Speaker Identification From Youtube Obtained Data | 7 pages, 5 figures, 1 Table, Signal & Image Processing : An
International Journal (SIPIJ) Vol.5, No.5, October 2014 | null | 10.5121/sipij.2014.5503 | null | cs.SD cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | An efficient, and intuitive algorithm is presented for the identification of
speakers from a long dataset (like YouTube long discussion, Cocktail party
recorded audio or video).The goal of automatic speaker identification is to
identify the number of different speakers and prepare a model for that speaker
by extraction, characterization and speaker-specific information contained in
the speech signal. It has many diverse application specially in the field of
Surveillance, Immigrations at Airport, cyber security, transcription in
multi-source of similar sound source, where it is difficult to assign
transcription arbitrary. The most commonly speech parametrization used in
speaker verification, K-mean, cepstral analysis, is detailed. Gaussian mixture
modeling, which is the speaker modeling technique is then explained. Gaussian
mixture models (GMM), perhaps the most robust machine learning algorithm has
been introduced examine and judge carefully speaker identification in text
independent. The application or employment of Gaussian mixture models for
monitoring & Analysing speaker identity is encouraged by the familiarity,
awareness, or understanding gained through experience that Gaussian spectrum
depict the characteristics of speaker's spectral conformational pattern and
remarkable ability of GMM to construct capricious densities after that we
illustrate 'Expectation maximization' an iterative algorithm which takes some
arbitrary value in initial estimation and carry on the iterative process until
the convergence of value is observed,so by doing various number of experiments
we are able to obtain 79 ~ 82% of identification rate using Vector quantization
and 85 ~ 92.6% of identification rate using GMM modeling by Expectation
maximization parameter estimation depending on variation of parameter.
| [
{
"version": "v1",
"created": "Tue, 11 Nov 2014 13:20:19 GMT"
}
] | 2014-11-12T00:00:00 | [
[
"Chaudhary",
"Nitesh Kumar",
""
]
] | TITLE: Speaker Identification From Youtube Obtained Data
ABSTRACT: An efficient, and intuitive algorithm is presented for the identification of
speakers from a long dataset (like YouTube long discussion, Cocktail party
recorded audio or video).The goal of automatic speaker identification is to
identify the number of different speakers and prepare a model for that speaker
by extraction, characterization and speaker-specific information contained in
the speech signal. It has many diverse application specially in the field of
Surveillance, Immigrations at Airport, cyber security, transcription in
multi-source of similar sound source, where it is difficult to assign
transcription arbitrary. The most commonly speech parametrization used in
speaker verification, K-mean, cepstral analysis, is detailed. Gaussian mixture
modeling, which is the speaker modeling technique is then explained. Gaussian
mixture models (GMM), perhaps the most robust machine learning algorithm has
been introduced examine and judge carefully speaker identification in text
independent. The application or employment of Gaussian mixture models for
monitoring & Analysing speaker identity is encouraged by the familiarity,
awareness, or understanding gained through experience that Gaussian spectrum
depict the characteristics of speaker's spectral conformational pattern and
remarkable ability of GMM to construct capricious densities after that we
illustrate 'Expectation maximization' an iterative algorithm which takes some
arbitrary value in initial estimation and carry on the iterative process until
the convergence of value is observed,so by doing various number of experiments
we are able to obtain 79 ~ 82% of identification rate using Vector quantization
and 85 ~ 92.6% of identification rate using GMM modeling by Expectation
maximization parameter estimation depending on variation of parameter.
|
1411.2821 | Saeed Afshar | Saeed Afshar, Libin George, Jonathan Tapson, Andre van Schaik, Philip
de Chazal, Tara Julia Hamilton | Turn Down that Noise: Synaptic Encoding of Afferent SNR in a Single
Spiking Neuron | null | null | null | null | cs.NE q-bio.NC | http://creativecommons.org/licenses/by/3.0/ | We have added a simplified neuromorphic model of Spike Time Dependent
Plasticity (STDP) to the Synapto-dendritic Kernel Adapting Neuron (SKAN). The
resulting neuron model is the first to show synaptic encoding of afferent
signal to noise ratio in addition to the unsupervised learning of spatio
temporal spike patterns. The neuron model is particularly suitable for
implementation in digital neuromorphic hardware as it does not use any complex
mathematical operations and uses a novel approach to achieve synaptic
homeostasis. The neurons noise compensation properties are characterized and
tested on noise corrupted zeros digits of the MNIST handwritten dataset.
Results show the simultaneously learning common patterns in its input data
while dynamically weighing individual afferent channels based on their signal
to noise ratio. Despite its simplicity the interesting behaviors of the neuron
model and the resulting computational power may offer insights into biological
systems.
| [
{
"version": "v1",
"created": "Tue, 11 Nov 2014 14:22:37 GMT"
}
] | 2014-11-12T00:00:00 | [
[
"Afshar",
"Saeed",
""
],
[
"George",
"Libin",
""
],
[
"Tapson",
"Jonathan",
""
],
[
"van Schaik",
"Andre",
""
],
[
"de Chazal",
"Philip",
""
],
[
"Hamilton",
"Tara Julia",
""
]
] | TITLE: Turn Down that Noise: Synaptic Encoding of Afferent SNR in a Single
Spiking Neuron
ABSTRACT: We have added a simplified neuromorphic model of Spike Time Dependent
Plasticity (STDP) to the Synapto-dendritic Kernel Adapting Neuron (SKAN). The
resulting neuron model is the first to show synaptic encoding of afferent
signal to noise ratio in addition to the unsupervised learning of spatio
temporal spike patterns. The neuron model is particularly suitable for
implementation in digital neuromorphic hardware as it does not use any complex
mathematical operations and uses a novel approach to achieve synaptic
homeostasis. The neurons noise compensation properties are characterized and
tested on noise corrupted zeros digits of the MNIST handwritten dataset.
Results show the simultaneously learning common patterns in its input data
while dynamically weighing individual afferent channels based on their signal
to noise ratio. Despite its simplicity the interesting behaviors of the neuron
model and the resulting computational power may offer insights into biological
systems.
|
1401.1456 | Margarita Karkali | Margarita Karkali, Francois Rousseau, Alexandros Ntoulas, Michalis
Vazirgiannis | Using temporal IDF for efficient novelty detection in text streams | 30 pages | null | null | null | cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Novelty detection in text streams is a challenging task that emerges in quite
a few different scenarios, ranging from email thread filtering to RSS news feed
recommendation on a smartphone. An efficient novelty detection algorithm can
save the user a great deal of time and resources when browsing through relevant
yet usually previously-seen content. Most of the recent research on detection
of novel documents in text streams has been building upon either geometric
distances or distributional similarities, with the former typically performing
better but being much slower due to the need of comparing an incoming document
with all the previously-seen ones. In this paper, we propose a new approach to
novelty detection in text streams. We describe a resource-aware mechanism that
is able to handle massive text streams such as the ones present today thanks to
the burst of social media and the emergence of the Web as the main source of
information. We capitalize on the historical Inverse Document Frequency (IDF)
that was known for capturing well term specificity and we show that it can be
used successfully at the document level as a measure of document novelty. This
enables us to avoid similarity comparisons with previous documents in the text
stream, thus scaling better and leading to faster execution times. Moreover, as
the collection of documents evolves over time, we use a temporal variant of IDF
not only to maintain an efficient representation of what has already been seen
but also to decay the document frequencies as the time goes by. We evaluate the
performance of the proposed approach on a real-world news articles dataset
created for this task. The results show that the proposed method outperforms
all of the baselines while managing to operate efficiently in terms of time
complexity and memory usage, which are of great importance in a mobile setting
scenario.
| [
{
"version": "v1",
"created": "Tue, 7 Jan 2014 17:43:37 GMT"
},
{
"version": "v2",
"created": "Sun, 9 Nov 2014 15:58:35 GMT"
}
] | 2014-11-11T00:00:00 | [
[
"Karkali",
"Margarita",
""
],
[
"Rousseau",
"Francois",
""
],
[
"Ntoulas",
"Alexandros",
""
],
[
"Vazirgiannis",
"Michalis",
""
]
] | TITLE: Using temporal IDF for efficient novelty detection in text streams
ABSTRACT: Novelty detection in text streams is a challenging task that emerges in quite
a few different scenarios, ranging from email thread filtering to RSS news feed
recommendation on a smartphone. An efficient novelty detection algorithm can
save the user a great deal of time and resources when browsing through relevant
yet usually previously-seen content. Most of the recent research on detection
of novel documents in text streams has been building upon either geometric
distances or distributional similarities, with the former typically performing
better but being much slower due to the need of comparing an incoming document
with all the previously-seen ones. In this paper, we propose a new approach to
novelty detection in text streams. We describe a resource-aware mechanism that
is able to handle massive text streams such as the ones present today thanks to
the burst of social media and the emergence of the Web as the main source of
information. We capitalize on the historical Inverse Document Frequency (IDF)
that was known for capturing well term specificity and we show that it can be
used successfully at the document level as a measure of document novelty. This
enables us to avoid similarity comparisons with previous documents in the text
stream, thus scaling better and leading to faster execution times. Moreover, as
the collection of documents evolves over time, we use a temporal variant of IDF
not only to maintain an efficient representation of what has already been seen
but also to decay the document frequencies as the time goes by. We evaluate the
performance of the proposed approach on a real-world news articles dataset
created for this task. The results show that the proposed method outperforms
all of the baselines while managing to operate efficiently in terms of time
complexity and memory usage, which are of great importance in a mobile setting
scenario.
|
1402.2648 | Deliang Fan Deliang Fan | Deliang Fan, Mrigank Sharad, Kaushik Roy | Design and Synthesis of Ultra Low Energy Spin-Memristor Threshold Logic | this paper is submitted to IEEE Transactions on Nanotechnology. It is
currently under review | IEEE Transactions on Nanotechnology, Volume:13, Issue: 3, 2014 | 10.1109/TNANO.2014.2312177 | null | cs.ET cond-mat.mes-hall | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A threshold logic gate (TLG) performs weighted sum of multiple inputs and
compares the sum with a threshold. We propose Spin-Memeristor Threshold Logic
(SMTL) gates, which employ memristive cross-bar array (MCA) to perform
current-mode summation of binary inputs, whereas, the low-voltage
fast-switching spintronic threshold devices (STD) carry out the threshold
operation in an energy efficient manner. Field programmable SMTL gate arrays
can operate at a small terminal voltage of ~50mV, resulting in ultra-low power
consumption in gates as well as programmable interconnect networks. We evaluate
the performance of SMTL using threshold logic synthesis. Results for common
benchmarks show that SMTL based programmable logic hardware can be more than
100x energy efficient than state of the art CMOS FPGA.
| [
{
"version": "v1",
"created": "Mon, 10 Feb 2014 19:59:29 GMT"
}
] | 2014-11-11T00:00:00 | [
[
"Fan",
"Deliang",
""
],
[
"Sharad",
"Mrigank",
""
],
[
"Roy",
"Kaushik",
""
]
] | TITLE: Design and Synthesis of Ultra Low Energy Spin-Memristor Threshold Logic
ABSTRACT: A threshold logic gate (TLG) performs weighted sum of multiple inputs and
compares the sum with a threshold. We propose Spin-Memeristor Threshold Logic
(SMTL) gates, which employ memristive cross-bar array (MCA) to perform
current-mode summation of binary inputs, whereas, the low-voltage
fast-switching spintronic threshold devices (STD) carry out the threshold
operation in an energy efficient manner. Field programmable SMTL gate arrays
can operate at a small terminal voltage of ~50mV, resulting in ultra-low power
consumption in gates as well as programmable interconnect networks. We evaluate
the performance of SMTL using threshold logic synthesis. Results for common
benchmarks show that SMTL based programmable logic hardware can be more than
100x energy efficient than state of the art CMOS FPGA.
|
1406.4607 | Camellia Sarkar | Sarika Jalan, Camellia Sarkar, Anagha Madhusudanan, Sanjiv Kumar
Dwivedi | Uncovering Randomness and Success in Society | 39 pages, 12 figures, 14 tables | PloS one, 9(2), e88249 (2014) | 10.1371/journal.pone.0088249 | null | physics.soc-ph cs.SI nlin.AO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | An understanding of how individuals shape and impact the evolution of society
is vastly limited due to the unavailability of large-scale reliable datasets
that can simultaneously capture information regarding individual movements and
social interactions. We believe that the popular Indian film industry,
'Bollywood', can provide a social network apt for such a study. Bollywood
provides massive amounts of real, unbiased data that spans more than 100 years,
and hence this network has been used as a model for the present paper. The
nodes which maintain a moderate degree or widely cooperate with the other nodes
of the network tend to be more fit (measured as the success of the node in the
industry) in comparison to the other nodes. The analysis carried forth in the
current work, using a conjoined framework of complex network theory and random
matrix theory, aims to quantify the elements that determine the fitness of an
individual node and the factors that contribute to the robustness of a network.
The authors of this paper believe that the method of study used in the current
paper can be extended to study various other industries and organizations.
| [
{
"version": "v1",
"created": "Wed, 18 Jun 2014 05:48:54 GMT"
},
{
"version": "v2",
"created": "Fri, 27 Jun 2014 08:21:32 GMT"
},
{
"version": "v3",
"created": "Tue, 8 Jul 2014 10:54:39 GMT"
}
] | 2014-11-11T00:00:00 | [
[
"Jalan",
"Sarika",
""
],
[
"Sarkar",
"Camellia",
""
],
[
"Madhusudanan",
"Anagha",
""
],
[
"Dwivedi",
"Sanjiv Kumar",
""
]
] | TITLE: Uncovering Randomness and Success in Society
ABSTRACT: An understanding of how individuals shape and impact the evolution of society
is vastly limited due to the unavailability of large-scale reliable datasets
that can simultaneously capture information regarding individual movements and
social interactions. We believe that the popular Indian film industry,
'Bollywood', can provide a social network apt for such a study. Bollywood
provides massive amounts of real, unbiased data that spans more than 100 years,
and hence this network has been used as a model for the present paper. The
nodes which maintain a moderate degree or widely cooperate with the other nodes
of the network tend to be more fit (measured as the success of the node in the
industry) in comparison to the other nodes. The analysis carried forth in the
current work, using a conjoined framework of complex network theory and random
matrix theory, aims to quantify the elements that determine the fitness of an
individual node and the factors that contribute to the robustness of a network.
The authors of this paper believe that the method of study used in the current
paper can be extended to study various other industries and organizations.
|
1409.4899 | Michael Schreiber | Michael Schreiber | Is the new citation-rank approach P100' in bibliometrics really new? | 11 pages, 4 figures, 5 tables | Journal of Informetrics 8, 997-1004 (2014) | 10.1016/j.joi.2014.10.001 | null | cs.DL | http://creativecommons.org/licenses/by-nc-sa/3.0/ | The percentile-based rating scale P100 describes the citation impact in terms
of the distribution of unique citation values. This approach has recently been
refined by considering also the frequency of papers with the same citation
counts. Here I compare the resulting P100' with P100 for an empirical dataset
and a simple fictitious model dataset. It is shown that P100' is not much
different from standard percentile-based ratings in terms of citation
frequencies. A new indicator P100'' is introduced.
| [
{
"version": "v1",
"created": "Wed, 17 Sep 2014 08:22:32 GMT"
}
] | 2014-11-11T00:00:00 | [
[
"Schreiber",
"Michael",
""
]
] | TITLE: Is the new citation-rank approach P100' in bibliometrics really new?
ABSTRACT: The percentile-based rating scale P100 describes the citation impact in terms
of the distribution of unique citation values. This approach has recently been
refined by considering also the frequency of papers with the same citation
counts. Here I compare the resulting P100' with P100 for an empirical dataset
and a simple fictitious model dataset. It is shown that P100' is not much
different from standard percentile-based ratings in terms of citation
frequencies. A new indicator P100'' is introduced.
|
1409.5908 | Nipun Batra | Jack Kelly, Nipun Batra, Oliver Parson, Haimonti Dutta, William
Knottenbelt, Alex Rogers, Amarjeet Singh, Mani Srivastava | Demo Abstract: NILMTK v0.2: A Non-intrusive Load Monitoring Toolkit for
Large Scale Data Sets | 1st ACM International Conference on Embedded Systems For
Energy-Efficient Buildings, 2014 | null | 10.1145/2674061.2675024 | null | cs.OH | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this demonstration, we present an open source toolkit for evaluating
non-intrusive load monitoring research; a field which aims to disaggregate a
household's total electricity consumption into individual appliances. The
toolkit contains: a number of importers for existing public data sets, a set of
preprocessing and statistics functions, a benchmark disaggregation algorithm
and a set of metrics to evaluate the performance of such algorithms.
Specifically, this release of the toolkit has been designed to enable the use
of large data sets by only loading individual chunks of the whole data set into
memory at once for processing, before combining the results of each chunk.
| [
{
"version": "v1",
"created": "Sat, 20 Sep 2014 17:36:43 GMT"
},
{
"version": "v2",
"created": "Wed, 8 Oct 2014 16:43:18 GMT"
},
{
"version": "v3",
"created": "Sun, 9 Nov 2014 20:43:51 GMT"
}
] | 2014-11-11T00:00:00 | [
[
"Kelly",
"Jack",
""
],
[
"Batra",
"Nipun",
""
],
[
"Parson",
"Oliver",
""
],
[
"Dutta",
"Haimonti",
""
],
[
"Knottenbelt",
"William",
""
],
[
"Rogers",
"Alex",
""
],
[
"Singh",
"Amarjeet",
""
],
[
"Srivastava",
"Mani",
""
]
] | TITLE: Demo Abstract: NILMTK v0.2: A Non-intrusive Load Monitoring Toolkit for
Large Scale Data Sets
ABSTRACT: In this demonstration, we present an open source toolkit for evaluating
non-intrusive load monitoring research; a field which aims to disaggregate a
household's total electricity consumption into individual appliances. The
toolkit contains: a number of importers for existing public data sets, a set of
preprocessing and statistics functions, a benchmark disaggregation algorithm
and a set of metrics to evaluate the performance of such algorithms.
Specifically, this release of the toolkit has been designed to enable the use
of large data sets by only loading individual chunks of the whole data set into
memory at once for processing, before combining the results of each chunk.
|
1411.2079 | Nick Gravin | Ning Chen, Nick Gravin, Pinyan Lu | Competitive analysis via benchmark decomposition | null | null | null | null | cs.GT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a uniform approach for the design and analysis of prior-free
competitive auctions and online auctions. Our philosophy is to view the
benchmark function as a variable parameter of the model and study a broad class
of functions instead of a individual target benchmark. We consider a multitude
of well-studied auction settings, and improve upon a few previous results.
(1) Multi-unit auctions. Given a $\beta$-competitive unlimited supply
auction, the best previously known multi-unit auction is $2\beta$-competitive.
We design a $(1+\beta)$-competitive auction reducing the ratio from $4.84$ to
$3.24$. These results carry over to matroid and position auctions.
(2) General downward-closed environments. We design a $6.5$-competitive
auction improving upon the ratio of $7.5$. Our auction is noticeably simpler
than the previous best one.
(3) Unlimited supply online auctions. Our analysis yields an auction with a
competitive ratio of $4.12$, which significantly narrows the margin of
$[4,4.84]$ previously known for this problem.
A particularly important tool in our analysis is a simple decomposition
lemma, which allows us to bound the competitive ratio against a sum of
benchmark functions. We use this lemma in a "divide and conquer" fashion by
dividing the target benchmark into the sum of simpler functions.
| [
{
"version": "v1",
"created": "Sat, 8 Nov 2014 04:30:45 GMT"
}
] | 2014-11-11T00:00:00 | [
[
"Chen",
"Ning",
""
],
[
"Gravin",
"Nick",
""
],
[
"Lu",
"Pinyan",
""
]
] | TITLE: Competitive analysis via benchmark decomposition
ABSTRACT: We propose a uniform approach for the design and analysis of prior-free
competitive auctions and online auctions. Our philosophy is to view the
benchmark function as a variable parameter of the model and study a broad class
of functions instead of a individual target benchmark. We consider a multitude
of well-studied auction settings, and improve upon a few previous results.
(1) Multi-unit auctions. Given a $\beta$-competitive unlimited supply
auction, the best previously known multi-unit auction is $2\beta$-competitive.
We design a $(1+\beta)$-competitive auction reducing the ratio from $4.84$ to
$3.24$. These results carry over to matroid and position auctions.
(2) General downward-closed environments. We design a $6.5$-competitive
auction improving upon the ratio of $7.5$. Our auction is noticeably simpler
than the previous best one.
(3) Unlimited supply online auctions. Our analysis yields an auction with a
competitive ratio of $4.12$, which significantly narrows the margin of
$[4,4.84]$ previously known for this problem.
A particularly important tool in our analysis is a simple decomposition
lemma, which allows us to bound the competitive ratio against a sum of
benchmark functions. We use this lemma in a "divide and conquer" fashion by
dividing the target benchmark into the sum of simpler functions.
|
1411.2153 | Simone Cirillo | Simone Cirillo, Stefan Lloyd, Peter Nordin | Evolving intraday foreign exchange trading strategies utilizing multiple
instruments price series | 15 pages, 10 figures, 9 tables | null | null | null | cs.NE q-fin.TR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a Genetic Programming architecture for the generation of foreign
exchange trading strategies. The system's principal features are the evolution
of free-form strategies which do not rely on any prior models and the
utilization of price series from multiple instruments as input data. This
latter feature constitutes an innovation with respect to previous works
documented in literature. In this article we utilize Open, High, Low, Close bar
data at a 5 minutes frequency for the AUD.USD, EUR.USD, GBP.USD and USD.JPY
currency pairs. We will test the implementation analyzing the in-sample and
out-of-sample performance of strategies for trading the USD.JPY obtained across
multiple algorithm runs. We will also evaluate the differences between
strategies selected according to two different criteria: one relies on the
fitness obtained on the training set only, the second one makes use of an
additional validation dataset. Strategy activity and trade accuracy are
remarkably stable between in and out of sample results. From a profitability
aspect, the two criteria both result in strategies successful on out-of-sample
data but exhibiting different characteristics. The overall best performing
out-of-sample strategy achieves a yearly return of 19%.
| [
{
"version": "v1",
"created": "Sat, 8 Nov 2014 19:22:55 GMT"
}
] | 2014-11-11T00:00:00 | [
[
"Cirillo",
"Simone",
""
],
[
"Lloyd",
"Stefan",
""
],
[
"Nordin",
"Peter",
""
]
] | TITLE: Evolving intraday foreign exchange trading strategies utilizing multiple
instruments price series
ABSTRACT: We propose a Genetic Programming architecture for the generation of foreign
exchange trading strategies. The system's principal features are the evolution
of free-form strategies which do not rely on any prior models and the
utilization of price series from multiple instruments as input data. This
latter feature constitutes an innovation with respect to previous works
documented in literature. In this article we utilize Open, High, Low, Close bar
data at a 5 minutes frequency for the AUD.USD, EUR.USD, GBP.USD and USD.JPY
currency pairs. We will test the implementation analyzing the in-sample and
out-of-sample performance of strategies for trading the USD.JPY obtained across
multiple algorithm runs. We will also evaluate the differences between
strategies selected according to two different criteria: one relies on the
fitness obtained on the training set only, the second one makes use of an
additional validation dataset. Strategy activity and trade accuracy are
remarkably stable between in and out of sample results. From a profitability
aspect, the two criteria both result in strategies successful on out-of-sample
data but exhibiting different characteristics. The overall best performing
out-of-sample strategy achieves a yearly return of 19%.
|
1411.2173 | Julieta Martinez | Julieta Martinez, Holger H. Hoos, James J. Little | Stacked Quantizers for Compositional Vector Compression | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently, Babenko and Lempitsky introduced Additive Quantization (AQ), a
generalization of Product Quantization (PQ) where a non-independent set of
codebooks is used to compress vectors into small binary codes. Unfortunately,
under this scheme encoding cannot be done independently in each codebook, and
optimal encoding is an NP-hard problem. In this paper, we observe that PQ and
AQ are both compositional quantizers that lie on the extremes of the codebook
dependence-independence assumption, and explore an intermediate approach that
exploits a hierarchical structure in the codebooks. This results in a method
that achieves quantization error on par with or lower than AQ, while being
several orders of magnitude faster. We perform a complexity analysis of PQ, AQ
and our method, and evaluate our approach on standard benchmarks of SIFT and
GIST descriptors, as well as on new datasets of features obtained from
state-of-the-art convolutional neural networks.
| [
{
"version": "v1",
"created": "Sat, 8 Nov 2014 22:51:12 GMT"
}
] | 2014-11-11T00:00:00 | [
[
"Martinez",
"Julieta",
""
],
[
"Hoos",
"Holger H.",
""
],
[
"Little",
"James J.",
""
]
] | TITLE: Stacked Quantizers for Compositional Vector Compression
ABSTRACT: Recently, Babenko and Lempitsky introduced Additive Quantization (AQ), a
generalization of Product Quantization (PQ) where a non-independent set of
codebooks is used to compress vectors into small binary codes. Unfortunately,
under this scheme encoding cannot be done independently in each codebook, and
optimal encoding is an NP-hard problem. In this paper, we observe that PQ and
AQ are both compositional quantizers that lie on the extremes of the codebook
dependence-independence assumption, and explore an intermediate approach that
exploits a hierarchical structure in the codebooks. This results in a method
that achieves quantization error on par with or lower than AQ, while being
several orders of magnitude faster. We perform a complexity analysis of PQ, AQ
and our method, and evaluate our approach on standard benchmarks of SIFT and
GIST descriptors, as well as on new datasets of features obtained from
state-of-the-art convolutional neural networks.
|
1411.2214 | Babak Saleh | Babak Saleh, Ali Farhadi, Ahmed Elgammal | Abnormal Object Recognition: A Comprehensive Study | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | When describing images, humans tend not to talk about the obvious, but rather
mention what they find interesting. We argue that abnormalities and deviations
from typicalities are among the most important components that form what is
worth mentioning. In this paper we introduce the abnormality detection as a
recognition problem and show how to model typicalities and, consequently,
meaningful deviations from prototypical properties of categories. Our model can
recognize abnormalities and report the main reasons of any recognized
abnormality. We introduce the abnormality detection dataset and show
interesting results on how to reason about abnormalities.
| [
{
"version": "v1",
"created": "Sun, 9 Nov 2014 09:51:06 GMT"
}
] | 2014-11-11T00:00:00 | [
[
"Saleh",
"Babak",
""
],
[
"Farhadi",
"Ali",
""
],
[
"Elgammal",
"Ahmed",
""
]
] | TITLE: Abnormal Object Recognition: A Comprehensive Study
ABSTRACT: When describing images, humans tend not to talk about the obvious, but rather
mention what they find interesting. We argue that abnormalities and deviations
from typicalities are among the most important components that form what is
worth mentioning. In this paper we introduce the abnormality detection as a
recognition problem and show how to model typicalities and, consequently,
meaningful deviations from prototypical properties of categories. Our model can
recognize abnormalities and report the main reasons of any recognized
abnormality. We introduce the abnormality detection dataset and show
interesting results on how to reason about abnormalities.
|
1411.2331 | Makoto Yamada | Makoto Yamada, Avishek Saha, Hua Ouyang, Dawei Yin, Yi Chang | N$^3$LARS: Minimum Redundancy Maximum Relevance Feature Selection for
Large and High-dimensional Data | arXiv admin note: text overlap with arXiv:1202.0515 | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a feature selection method that finds non-redundant features from
a large and high-dimensional data in nonlinear way. Specifically, we propose a
nonlinear extension of the non-negative least-angle regression (LARS) called
N${}^3$LARS, where the similarity between input and output is measured through
the normalized version of the Hilbert-Schmidt Independence Criterion (HSIC). An
advantage of N${}^3$LARS is that it can easily incorporate with map-reduce
frameworks such as Hadoop and Spark. Thus, with the help of distributed
computing, a set of features can be efficiently selected from a large and
high-dimensional data. Moreover, N${}^3$LARS is a convex method and can find a
global optimum solution. The effectiveness of the proposed method is first
demonstrated through feature selection experiments for classification and
regression with small and high-dimensional datasets. Finally, we evaluate our
proposed method over a large and high-dimensional biology dataset.
| [
{
"version": "v1",
"created": "Mon, 10 Nov 2014 05:43:28 GMT"
}
] | 2014-11-11T00:00:00 | [
[
"Yamada",
"Makoto",
""
],
[
"Saha",
"Avishek",
""
],
[
"Ouyang",
"Hua",
""
],
[
"Yin",
"Dawei",
""
],
[
"Chang",
"Yi",
""
]
] | TITLE: N$^3$LARS: Minimum Redundancy Maximum Relevance Feature Selection for
Large and High-dimensional Data
ABSTRACT: We propose a feature selection method that finds non-redundant features from
a large and high-dimensional data in nonlinear way. Specifically, we propose a
nonlinear extension of the non-negative least-angle regression (LARS) called
N${}^3$LARS, where the similarity between input and output is measured through
the normalized version of the Hilbert-Schmidt Independence Criterion (HSIC). An
advantage of N${}^3$LARS is that it can easily incorporate with map-reduce
frameworks such as Hadoop and Spark. Thus, with the help of distributed
computing, a set of features can be efficiently selected from a large and
high-dimensional data. Moreover, N${}^3$LARS is a convex method and can find a
global optimum solution. The effectiveness of the proposed method is first
demonstrated through feature selection experiments for classification and
regression with small and high-dimensional datasets. Finally, we evaluate our
proposed method over a large and high-dimensional biology dataset.
|
1411.2377 | Tomonori Kouya | Tomonori Kouya | A Highly Efficient Implementation of Multiple Precision Sparse
Matrix-Vector Multiplication and Its Application to Product-type Krylov
Subspace Methods | null | International Journal of Numerical Methods and Applications,
Vol.7, Issue 2, 2012, Pages 107 - 119 | null | null | math.NA cs.NA | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We evaluate the performance of the Krylov subspace method by using highly
efficient multiple precision sparse matrix-vector multiplication (SpMV).
BNCpack is our multiple precision numerical computation library based on
MPFR/GMP, which is one of the most efficient arbitrary precision floating-point
arithmetic libraries. However, it does not include functions that can
manipulate multiple precision sparse matrices. Therefore, by using benchmark
tests, we show that SpMV implemented in these functions can be more efficient.
Finally, we also show that product-type Krylov subspace methods such as BiCG
and GPBiCG in which we have embedded SpMV, can efficiently solve large-scale
linear systems of equations provided in the UF sparse matrix collections in a
memory-restricted computing environment.
| [
{
"version": "v1",
"created": "Mon, 10 Nov 2014 10:58:18 GMT"
}
] | 2014-11-11T00:00:00 | [
[
"Kouya",
"Tomonori",
""
]
] | TITLE: A Highly Efficient Implementation of Multiple Precision Sparse
Matrix-Vector Multiplication and Its Application to Product-type Krylov
Subspace Methods
ABSTRACT: We evaluate the performance of the Krylov subspace method by using highly
efficient multiple precision sparse matrix-vector multiplication (SpMV).
BNCpack is our multiple precision numerical computation library based on
MPFR/GMP, which is one of the most efficient arbitrary precision floating-point
arithmetic libraries. However, it does not include functions that can
manipulate multiple precision sparse matrices. Therefore, by using benchmark
tests, we show that SpMV implemented in these functions can be more efficient.
Finally, we also show that product-type Krylov subspace methods such as BiCG
and GPBiCG in which we have embedded SpMV, can efficiently solve large-scale
linear systems of equations provided in the UF sparse matrix collections in a
memory-restricted computing environment.
|
1411.2553 | Jeremy Morse | Jeremy Morse | Measuring the impact of input data on energy consumption of software | null | null | null | null | cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The amount of energy consumed during the execution of software, and the
ability to predict future consumption, is an important factor in the design of
embedded electronic systems. In this technical report I examine factors in the
execution of software that can affect energy consumption. Taking a simple
embedded software benchmark I measure to what extent input data can affect
energy consumption, and propose a method for reflecting this in software energy
models.
| [
{
"version": "v1",
"created": "Mon, 10 Nov 2014 19:42:34 GMT"
}
] | 2014-11-11T00:00:00 | [
[
"Morse",
"Jeremy",
""
]
] | TITLE: Measuring the impact of input data on energy consumption of software
ABSTRACT: The amount of energy consumed during the execution of software, and the
ability to predict future consumption, is an important factor in the design of
embedded electronic systems. In this technical report I examine factors in the
execution of software that can affect energy consumption. Taking a simple
embedded software benchmark I measure to what extent input data can affect
energy consumption, and propose a method for reflecting this in software energy
models.
|
1402.3902 | Karthikeyan Shanmugam | Murat Kocaoglu, Karthikeyan Shanmugam, Alexandros G. Dimakis and Adam
Klivans | Sparse Polynomial Learning and Graph Sketching | 14 pages; to appear in NIPS 2014l Updated proof of Theorem 5 and some
other minor changes during revision | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Let $f:\{-1,1\}^n$ be a polynomial with at most $s$ non-zero real
coefficients. We give an algorithm for exactly reconstructing f given random
examples from the uniform distribution on $\{-1,1\}^n$ that runs in time
polynomial in $n$ and $2s$ and succeeds if the function satisfies the unique
sign property: there is one output value which corresponds to a unique set of
values of the participating parities. This sufficient condition is satisfied
when every coefficient of f is perturbed by a small random noise, or satisfied
with high probability when s parity functions are chosen randomly or when all
the coefficients are positive. Learning sparse polynomials over the Boolean
domain in time polynomial in $n$ and $2s$ is considered notoriously hard in the
worst-case. Our result shows that the problem is tractable for almost all
sparse polynomials. Then, we show an application of this result to hypergraph
sketching which is the problem of learning a sparse (both in the number of
hyperedges and the size of the hyperedges) hypergraph from uniformly drawn
random cuts. We also provide experimental results on a real world dataset.
| [
{
"version": "v1",
"created": "Mon, 17 Feb 2014 06:00:16 GMT"
},
{
"version": "v2",
"created": "Tue, 18 Feb 2014 06:56:27 GMT"
},
{
"version": "v3",
"created": "Wed, 5 Nov 2014 22:35:40 GMT"
},
{
"version": "v4",
"created": "Fri, 7 Nov 2014 03:00:28 GMT"
}
] | 2014-11-10T00:00:00 | [
[
"Kocaoglu",
"Murat",
""
],
[
"Shanmugam",
"Karthikeyan",
""
],
[
"Dimakis",
"Alexandros G.",
""
],
[
"Klivans",
"Adam",
""
]
] | TITLE: Sparse Polynomial Learning and Graph Sketching
ABSTRACT: Let $f:\{-1,1\}^n$ be a polynomial with at most $s$ non-zero real
coefficients. We give an algorithm for exactly reconstructing f given random
examples from the uniform distribution on $\{-1,1\}^n$ that runs in time
polynomial in $n$ and $2s$ and succeeds if the function satisfies the unique
sign property: there is one output value which corresponds to a unique set of
values of the participating parities. This sufficient condition is satisfied
when every coefficient of f is perturbed by a small random noise, or satisfied
with high probability when s parity functions are chosen randomly or when all
the coefficients are positive. Learning sparse polynomials over the Boolean
domain in time polynomial in $n$ and $2s$ is considered notoriously hard in the
worst-case. Our result shows that the problem is tractable for almost all
sparse polynomials. Then, we show an application of this result to hypergraph
sketching which is the problem of learning a sparse (both in the number of
hyperedges and the size of the hyperedges) hypergraph from uniformly drawn
random cuts. We also provide experimental results on a real world dataset.
|
1402.7015 | Fabian Pedregosa | Fabian Pedregosa (INRIA Saclay - Ile de France, INRIA Paris -
Rocquencourt), Michael Eickenberg (INRIA Saclay - Ile de France, LNAO),
Philippe Ciuciu (INRIA Saclay - Ile de France, NEUROSPIN), Bertrand Thirion
(INRIA Saclay - Ile de France, NEUROSPIN), Alexandre Gramfort (LTCI) | Data-driven HRF estimation for encoding and decoding models | appears in NeuroImage (2015) | null | 10.1016/j.neuroimage.2014.09.060 | null | cs.CE cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Despite the common usage of a canonical, data-independent, hemodynamic
response function (HRF), it is known that the shape of the HRF varies across
brain regions and subjects. This suggests that a data-driven estimation of this
function could lead to more statistical power when modeling BOLD fMRI data.
However, unconstrained estimation of the HRF can yield highly unstable results
when the number of free parameters is large. We develop a method for the joint
estimation of activation and HRF using a rank constraint causing the estimated
HRF to be equal across events/conditions, yet permitting it to be different
across voxels. Model estimation leads to an optimization problem that we
propose to solve with an efficient quasi-Newton method exploiting fast gradient
computations. This model, called GLM with Rank-1 constraint (R1-GLM), can be
extended to the setting of GLM with separate designs which has been shown to
improve decoding accuracy in brain activity decoding experiments. We compare 10
different HRF modeling methods in terms of encoding and decoding score in two
different datasets. Our results show that the R1-GLM model significantly
outperforms competing methods in both encoding and decoding settings,
positioning it as an attractive method both from the points of view of accuracy
and computational efficiency.
| [
{
"version": "v1",
"created": "Thu, 27 Feb 2014 18:50:58 GMT"
},
{
"version": "v2",
"created": "Sun, 6 Apr 2014 06:11:17 GMT"
},
{
"version": "v3",
"created": "Tue, 15 Jul 2014 11:14:00 GMT"
},
{
"version": "v4",
"created": "Mon, 6 Oct 2014 16:39:55 GMT"
},
{
"version": "v5",
"created": "Fri, 31 Oct 2014 13:47:01 GMT"
},
{
"version": "v6",
"created": "Fri, 7 Nov 2014 11:27:19 GMT"
}
] | 2014-11-10T00:00:00 | [
[
"Pedregosa",
"Fabian",
"",
"INRIA Saclay - Ile de France, INRIA Paris -\n Rocquencourt"
],
[
"Eickenberg",
"Michael",
"",
"INRIA Saclay - Ile de France, LNAO"
],
[
"Ciuciu",
"Philippe",
"",
"INRIA Saclay - Ile de France, NEUROSPIN"
],
[
"Thirion",
"Bertrand",
"",
"INRIA Saclay - Ile de France, NEUROSPIN"
],
[
"Gramfort",
"Alexandre",
"",
"LTCI"
]
] | TITLE: Data-driven HRF estimation for encoding and decoding models
ABSTRACT: Despite the common usage of a canonical, data-independent, hemodynamic
response function (HRF), it is known that the shape of the HRF varies across
brain regions and subjects. This suggests that a data-driven estimation of this
function could lead to more statistical power when modeling BOLD fMRI data.
However, unconstrained estimation of the HRF can yield highly unstable results
when the number of free parameters is large. We develop a method for the joint
estimation of activation and HRF using a rank constraint causing the estimated
HRF to be equal across events/conditions, yet permitting it to be different
across voxels. Model estimation leads to an optimization problem that we
propose to solve with an efficient quasi-Newton method exploiting fast gradient
computations. This model, called GLM with Rank-1 constraint (R1-GLM), can be
extended to the setting of GLM with separate designs which has been shown to
improve decoding accuracy in brain activity decoding experiments. We compare 10
different HRF modeling methods in terms of encoding and decoding score in two
different datasets. Our results show that the R1-GLM model significantly
outperforms competing methods in both encoding and decoding settings,
positioning it as an attractive method both from the points of view of accuracy
and computational efficiency.
|
1409.4297 | Sayyed Ali Mirsoleimani | S. Ali Mirsoleimani, Aske Plaat, Jos Vermaseren, and Jaap van den
Herik | Performance analysis of a 240 thread tournament level MCTS Go program on
the Intel Xeon Phi | 7 pages, 12 figues | null | null | null | cs.PF | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In 2013 Intel introduced the Xeon Phi, a new parallel co-processor board. The
Xeon Phi is a cache-coherent many-core shared memory architecture claiming
CPU-like versatility, programmability, high performance, and power efficiency.
The first published micro-benchmark studies indicate that many of Intel's
claims appear to be true. The current paper is the first study on the Phi of a
complex artificial intelligence application. It contains an open source MCTS
application for playing tournament quality Go (an oriental board game). We
report the first speedup figures for up to 240 parallel threads on a real
machine, allowing a direct comparison to previous simulation studies. After a
substantial amount of work, we observed that performance scales well up to 32
threads, largely confirming previous simulation results of this Go program,
although the performance surprisingly deteriorates between 32 and 240 threads.
Furthermore, we report (1) unexpected performance anomalies between the Xeon
Phi and Xeon CPU for small problem sizes and small numbers of threads, and (2)
that performance is sensitive to scheduling choices. Achieving good performance
on the Xeon Phi for complex programs is not straightforward; it requires a deep
understanding of (1) search patterns, (2) of scheduling, and (3) of the
architecture and its many cores and caches. In practice, the Xeon Phi is less
straightforward to program for than originally envisioned by Intel.
| [
{
"version": "v1",
"created": "Mon, 15 Sep 2014 15:37:11 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Nov 2014 21:18:03 GMT"
}
] | 2014-11-10T00:00:00 | [
[
"Mirsoleimani",
"S. Ali",
""
],
[
"Plaat",
"Aske",
""
],
[
"Vermaseren",
"Jos",
""
],
[
"Herik",
"Jaap van den",
""
]
] | TITLE: Performance analysis of a 240 thread tournament level MCTS Go program on
the Intel Xeon Phi
ABSTRACT: In 2013 Intel introduced the Xeon Phi, a new parallel co-processor board. The
Xeon Phi is a cache-coherent many-core shared memory architecture claiming
CPU-like versatility, programmability, high performance, and power efficiency.
The first published micro-benchmark studies indicate that many of Intel's
claims appear to be true. The current paper is the first study on the Phi of a
complex artificial intelligence application. It contains an open source MCTS
application for playing tournament quality Go (an oriental board game). We
report the first speedup figures for up to 240 parallel threads on a real
machine, allowing a direct comparison to previous simulation studies. After a
substantial amount of work, we observed that performance scales well up to 32
threads, largely confirming previous simulation results of this Go program,
although the performance surprisingly deteriorates between 32 and 240 threads.
Furthermore, we report (1) unexpected performance anomalies between the Xeon
Phi and Xeon CPU for small problem sizes and small numbers of threads, and (2)
that performance is sensitive to scheduling choices. Achieving good performance
on the Xeon Phi for complex programs is not straightforward; it requires a deep
understanding of (1) search patterns, (2) of scheduling, and (3) of the
architecture and its many cores and caches. In practice, the Xeon Phi is less
straightforward to program for than originally envisioned by Intel.
|
1411.1951 | Manuel P\"oter | Manuel P\"oter | Pheet meets C++11 | null | null | null | null | cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Pheet is a C++ task-scheduling framework that allows for easy customization
of internal data-structures. The implementation was started before the C++11
standard was committed and therefore did not use the new standardized memory
model but compiler/platform specific intrinsics for atomic memory operations.
This not only makes the implementation harder to port to other compilers or
architectures but also suffers from the fact that prior C++ versions did not
specify any memory model.
In this report I discuss the porting of one of the internal Pheet data
structures to the new memory model and provide reasoning about the correctness
based on the semantics of the memory consistency model. Using two benchmarks
from the Pheet benchmark suite I compare the performance of the original
against the new implementation which shows a significant speedup under certain
conditions on one of the two test machines.
| [
{
"version": "v1",
"created": "Fri, 7 Nov 2014 15:40:48 GMT"
}
] | 2014-11-10T00:00:00 | [
[
"Pöter",
"Manuel",
""
]
] | TITLE: Pheet meets C++11
ABSTRACT: Pheet is a C++ task-scheduling framework that allows for easy customization
of internal data-structures. The implementation was started before the C++11
standard was committed and therefore did not use the new standardized memory
model but compiler/platform specific intrinsics for atomic memory operations.
This not only makes the implementation harder to port to other compilers or
architectures but also suffers from the fact that prior C++ versions did not
specify any memory model.
In this report I discuss the porting of one of the internal Pheet data
structures to the new memory model and provide reasoning about the correctness
based on the semantics of the memory consistency model. Using two benchmarks
from the Pheet benchmark suite I compare the performance of the original
against the new implementation which shows a significant speedup under certain
conditions on one of the two test machines.
|
1406.2035 | Dani Yogatama | Dani Yogatama and Manaal Faruqui and Chris Dyer and Noah A. Smith | Learning Word Representations with Hierarchical Sparse Coding | null | null | null | null | cs.CL cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a new method for learning word representations using hierarchical
regularization in sparse coding inspired by the linguistic study of word
meanings. We show an efficient learning algorithm based on stochastic proximal
methods that is significantly faster than previous approaches, making it
possible to perform hierarchical sparse coding on a corpus of billions of word
tokens. Experiments on various benchmark tasks---word similarity ranking,
analogies, sentence completion, and sentiment analysis---demonstrate that the
method outperforms or is competitive with state-of-the-art methods. Our word
representations are available at
\url{http://www.ark.cs.cmu.edu/dyogatam/wordvecs/}.
| [
{
"version": "v1",
"created": "Sun, 8 Jun 2014 22:35:09 GMT"
},
{
"version": "v2",
"created": "Thu, 6 Nov 2014 14:26:21 GMT"
}
] | 2014-11-07T00:00:00 | [
[
"Yogatama",
"Dani",
""
],
[
"Faruqui",
"Manaal",
""
],
[
"Dyer",
"Chris",
""
],
[
"Smith",
"Noah A.",
""
]
] | TITLE: Learning Word Representations with Hierarchical Sparse Coding
ABSTRACT: We propose a new method for learning word representations using hierarchical
regularization in sparse coding inspired by the linguistic study of word
meanings. We show an efficient learning algorithm based on stochastic proximal
methods that is significantly faster than previous approaches, making it
possible to perform hierarchical sparse coding on a corpus of billions of word
tokens. Experiments on various benchmark tasks---word similarity ranking,
analogies, sentence completion, and sentiment analysis---demonstrate that the
method outperforms or is competitive with state-of-the-art methods. Our word
representations are available at
\url{http://www.ark.cs.cmu.edu/dyogatam/wordvecs/}.
|
1411.1509 | Zetao Chen | Zetao Chen, Obadiah Lam, Adam Jacobson and Michael Milford | Convolutional Neural Network-based Place Recognition | 8 pages, 11 figures, this paper has been accepted by 2014
Australasian Conference on Robotics and Automation (ACRA 2014) to be held in
University of Melbourne, Dec 2~4 | null | null | null | cs.CV cs.LG cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently Convolutional Neural Networks (CNNs) have been shown to achieve
state-of-the-art performance on various classification tasks. In this paper, we
present for the first time a place recognition technique based on CNN models,
by combining the powerful features learnt by CNNs with a spatial and sequential
filter. Applying the system to a 70 km benchmark place recognition dataset we
achieve a 75% increase in recall at 100% precision, significantly outperforming
all previous state of the art techniques. We also conduct a comprehensive
performance comparison of the utility of features from all 21 layers for place
recognition, both for the benchmark dataset and for a second dataset with more
significant viewpoint changes.
| [
{
"version": "v1",
"created": "Thu, 6 Nov 2014 07:03:15 GMT"
}
] | 2014-11-07T00:00:00 | [
[
"Chen",
"Zetao",
""
],
[
"Lam",
"Obadiah",
""
],
[
"Jacobson",
"Adam",
""
],
[
"Milford",
"Michael",
""
]
] | TITLE: Convolutional Neural Network-based Place Recognition
ABSTRACT: Recently Convolutional Neural Networks (CNNs) have been shown to achieve
state-of-the-art performance on various classification tasks. In this paper, we
present for the first time a place recognition technique based on CNN models,
by combining the powerful features learnt by CNNs with a spatial and sequential
filter. Applying the system to a 70 km benchmark place recognition dataset we
achieve a 75% increase in recall at 100% precision, significantly outperforming
all previous state of the art techniques. We also conduct a comprehensive
performance comparison of the utility of features from all 21 layers for place
recognition, both for the benchmark dataset and for a second dataset with more
significant viewpoint changes.
|
1411.1623 | Siddharth Sigtia | Siddharth Sigtia, Emmanouil Benetos, Nicolas Boulanger-Lewandowski,
Tillman Weyde, Artur S. d'Avila Garcez, Simon Dixon | A Hybrid Recurrent Neural Network For Music Transcription | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We investigate the problem of incorporating higher-level symbolic score-like
information into Automatic Music Transcription (AMT) systems to improve their
performance. We use recurrent neural networks (RNNs) and their variants as
music language models (MLMs) and present a generative architecture for
combining these models with predictions from a frame level acoustic classifier.
We also compare different neural network architectures for acoustic modeling.
The proposed model computes a distribution over possible output sequences given
the acoustic input signal and we present an algorithm for performing a global
search for good candidate transcriptions. The performance of the proposed model
is evaluated on piano music from the MAPS dataset and we observe that the
proposed model consistently outperforms existing transcription methods.
| [
{
"version": "v1",
"created": "Thu, 6 Nov 2014 14:18:39 GMT"
}
] | 2014-11-07T00:00:00 | [
[
"Sigtia",
"Siddharth",
""
],
[
"Benetos",
"Emmanouil",
""
],
[
"Boulanger-Lewandowski",
"Nicolas",
""
],
[
"Weyde",
"Tillman",
""
],
[
"Garcez",
"Artur S. d'Avila",
""
],
[
"Dixon",
"Simon",
""
]
] | TITLE: A Hybrid Recurrent Neural Network For Music Transcription
ABSTRACT: We investigate the problem of incorporating higher-level symbolic score-like
information into Automatic Music Transcription (AMT) systems to improve their
performance. We use recurrent neural networks (RNNs) and their variants as
music language models (MLMs) and present a generative architecture for
combining these models with predictions from a frame level acoustic classifier.
We also compare different neural network architectures for acoustic modeling.
The proposed model computes a distribution over possible output sequences given
the acoustic input signal and we present an algorithm for performing a global
search for good candidate transcriptions. The performance of the proposed model
is evaluated on piano music from the MAPS dataset and we observe that the
proposed model consistently outperforms existing transcription methods.
|
1411.1705 | Yuanyi Xue | Yuanyi Xue and Beril Erkin and Yao Wang | A Novel No-reference Video Quality Metric for Evaluating Temporal
Jerkiness due to Frame Freezing | null | null | 10.1109/TMM.2014.2368272 | null | cs.MM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we propose a novel no-reference (NR) video quality metric that
evaluates the impact of frame freezing due to either packet loss or late
arrival. Our metric uses a trained neural network acting on features that are
chosen to capture the impact of frame freezing on the perceived quality. The
considered features include the number of freezes, freeze duration statistics,
inter-freeze distance statistics, frame difference before and after the freeze,
normal frame difference, and the ratio of them. We use the neural network to
find the mapping between features and subjective test scores. We optimize the
network structure and the feature selection through a cross validation
procedure, using training samples extracted from both VQEG and LIVE video
databases. The resulting feature set and network structure yields accurate
quality prediction for both the training data containing 54 test videos and a
separate testing dataset including 14 videos, with Pearson Correlation
Coefficients greater than 0.9 and 0.8 for the training set and the testing set,
respectively. Our proposed metric has low complexity and could be utilized in a
system with realtime processing constraint.
| [
{
"version": "v1",
"created": "Wed, 5 Nov 2014 16:29:30 GMT"
}
] | 2014-11-07T00:00:00 | [
[
"Xue",
"Yuanyi",
""
],
[
"Erkin",
"Beril",
""
],
[
"Wang",
"Yao",
""
]
] | TITLE: A Novel No-reference Video Quality Metric for Evaluating Temporal
Jerkiness due to Frame Freezing
ABSTRACT: In this work, we propose a novel no-reference (NR) video quality metric that
evaluates the impact of frame freezing due to either packet loss or late
arrival. Our metric uses a trained neural network acting on features that are
chosen to capture the impact of frame freezing on the perceived quality. The
considered features include the number of freezes, freeze duration statistics,
inter-freeze distance statistics, frame difference before and after the freeze,
normal frame difference, and the ratio of them. We use the neural network to
find the mapping between features and subjective test scores. We optimize the
network structure and the feature selection through a cross validation
procedure, using training samples extracted from both VQEG and LIVE video
databases. The resulting feature set and network structure yields accurate
quality prediction for both the training data containing 54 test videos and a
separate testing dataset including 14 videos, with Pearson Correlation
Coefficients greater than 0.9 and 0.8 for the training set and the testing set,
respectively. Our proposed metric has low complexity and could be utilized in a
system with realtime processing constraint.
|
1404.7584 | Jo\~ao F. Henriques | Jo\~ao F. Henriques, Rui Caseiro, Pedro Martins, Jorge Batista | High-Speed Tracking with Kernelized Correlation Filters | null | null | 10.1109/TPAMI.2014.2345390 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The core component of most modern trackers is a discriminative classifier,
tasked with distinguishing between the target and the surrounding environment.
To cope with natural image changes, this classifier is typically trained with
translated and scaled sample patches. Such sets of samples are riddled with
redundancies -- any overlapping pixels are constrained to be the same. Based on
this simple observation, we propose an analytic model for datasets of thousands
of translated patches. By showing that the resulting data matrix is circulant,
we can diagonalize it with the Discrete Fourier Transform, reducing both
storage and computation by several orders of magnitude. Interestingly, for
linear regression our formulation is equivalent to a correlation filter, used
by some of the fastest competitive trackers. For kernel regression, however, we
derive a new Kernelized Correlation Filter (KCF), that unlike other kernel
algorithms has the exact same complexity as its linear counterpart. Building on
it, we also propose a fast multi-channel extension of linear correlation
filters, via a linear kernel, which we call Dual Correlation Filter (DCF). Both
KCF and DCF outperform top-ranking trackers such as Struck or TLD on a 50
videos benchmark, despite running at hundreds of frames-per-second, and being
implemented in a few lines of code (Algorithm 1). To encourage further
developments, our tracking framework was made open-source.
| [
{
"version": "v1",
"created": "Wed, 30 Apr 2014 04:16:38 GMT"
},
{
"version": "v2",
"created": "Fri, 11 Jul 2014 23:04:01 GMT"
},
{
"version": "v3",
"created": "Wed, 5 Nov 2014 01:32:56 GMT"
}
] | 2014-11-06T00:00:00 | [
[
"Henriques",
"João F.",
""
],
[
"Caseiro",
"Rui",
""
],
[
"Martins",
"Pedro",
""
],
[
"Batista",
"Jorge",
""
]
] | TITLE: High-Speed Tracking with Kernelized Correlation Filters
ABSTRACT: The core component of most modern trackers is a discriminative classifier,
tasked with distinguishing between the target and the surrounding environment.
To cope with natural image changes, this classifier is typically trained with
translated and scaled sample patches. Such sets of samples are riddled with
redundancies -- any overlapping pixels are constrained to be the same. Based on
this simple observation, we propose an analytic model for datasets of thousands
of translated patches. By showing that the resulting data matrix is circulant,
we can diagonalize it with the Discrete Fourier Transform, reducing both
storage and computation by several orders of magnitude. Interestingly, for
linear regression our formulation is equivalent to a correlation filter, used
by some of the fastest competitive trackers. For kernel regression, however, we
derive a new Kernelized Correlation Filter (KCF), that unlike other kernel
algorithms has the exact same complexity as its linear counterpart. Building on
it, we also propose a fast multi-channel extension of linear correlation
filters, via a linear kernel, which we call Dual Correlation Filter (DCF). Both
KCF and DCF outperform top-ranking trackers such as Struck or TLD on a 50
videos benchmark, despite running at hundreds of frames-per-second, and being
implemented in a few lines of code (Algorithm 1). To encourage further
developments, our tracking framework was made open-source.
|
1405.3531 | Ken Chatfield | Ken Chatfield, Karen Simonyan, Andrea Vedaldi, Andrew Zisserman | Return of the Devil in the Details: Delving Deep into Convolutional Nets | Published in proceedings of BMVC 2014 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The latest generation of Convolutional Neural Networks (CNN) have achieved
impressive results in challenging benchmarks on image recognition and object
detection, significantly raising the interest of the community in these
methods. Nevertheless, it is still unclear how different CNN methods compare
with each other and with previous state-of-the-art shallow representations such
as the Bag-of-Visual-Words and the Improved Fisher Vector. This paper conducts
a rigorous evaluation of these new techniques, exploring different deep
architectures and comparing them on a common ground, identifying and disclosing
important implementation details. We identify several useful properties of
CNN-based representations, including the fact that the dimensionality of the
CNN output layer can be reduced significantly without having an adverse effect
on performance. We also identify aspects of deep and shallow methods that can
be successfully shared. In particular, we show that the data augmentation
techniques commonly applied to CNN-based methods can also be applied to shallow
methods, and result in an analogous performance boost. Source code and models
to reproduce the experiments in the paper is made publicly available.
| [
{
"version": "v1",
"created": "Wed, 14 May 2014 15:19:22 GMT"
},
{
"version": "v2",
"created": "Thu, 15 May 2014 19:48:38 GMT"
},
{
"version": "v3",
"created": "Tue, 22 Jul 2014 15:09:39 GMT"
},
{
"version": "v4",
"created": "Wed, 5 Nov 2014 08:34:48 GMT"
}
] | 2014-11-06T00:00:00 | [
[
"Chatfield",
"Ken",
""
],
[
"Simonyan",
"Karen",
""
],
[
"Vedaldi",
"Andrea",
""
],
[
"Zisserman",
"Andrew",
""
]
] | TITLE: Return of the Devil in the Details: Delving Deep into Convolutional Nets
ABSTRACT: The latest generation of Convolutional Neural Networks (CNN) have achieved
impressive results in challenging benchmarks on image recognition and object
detection, significantly raising the interest of the community in these
methods. Nevertheless, it is still unclear how different CNN methods compare
with each other and with previous state-of-the-art shallow representations such
as the Bag-of-Visual-Words and the Improved Fisher Vector. This paper conducts
a rigorous evaluation of these new techniques, exploring different deep
architectures and comparing them on a common ground, identifying and disclosing
important implementation details. We identify several useful properties of
CNN-based representations, including the fact that the dimensionality of the
CNN output layer can be reduced significantly without having an adverse effect
on performance. We also identify aspects of deep and shallow methods that can
be successfully shared. In particular, we show that the data augmentation
techniques commonly applied to CNN-based methods can also be applied to shallow
methods, and result in an analogous performance boost. Source code and models
to reproduce the experiments in the paper is made publicly available.
|
1410.4485 | Miguel \'Angel Bautista Martin | Miguel \'Angel Bautista, Antonio Hern\'andez-Vela, Sergio Escalera,
Laura Igual, Oriol Pujol, Josep Moya, Ver\'onica Violant, Mar\'ia Teresa
Anguera | A Gesture Recognition System for Detecting Behavioral Patterns of ADHD | 12 pages | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present an application of gesture recognition using an extension of
Dynamic Time Warping (DTW) to recognize behavioural patterns of Attention
Deficit Hyperactivity Disorder (ADHD). We propose an extension of DTW using
one-class classifiers in order to be able to encode the variability of a
gesture category, and thus, perform an alignment between a gesture sample and a
gesture class. We model the set of gesture samples of a certain gesture
category using either GMMs or an approximation of Convex Hulls. Thus, we add a
theoretical contribution to classical warping path in DTW by including local
modeling of intra-class gesture variability. This methodology is applied in a
clinical context, detecting a group of ADHD behavioural patterns defined by
experts in psychology/psychiatry, to provide support to clinicians in the
diagnose procedure. The proposed methodology is tested on a novel multi-modal
dataset (RGB plus Depth) of ADHD children recordings with behavioural patterns.
We obtain satisfying results when compared to standard state-of-the-art
approaches in the DTW context.
| [
{
"version": "v1",
"created": "Thu, 16 Oct 2014 16:25:29 GMT"
},
{
"version": "v2",
"created": "Wed, 5 Nov 2014 10:25:13 GMT"
}
] | 2014-11-06T00:00:00 | [
[
"Bautista",
"Miguel Ángel",
""
],
[
"Hernández-Vela",
"Antonio",
""
],
[
"Escalera",
"Sergio",
""
],
[
"Igual",
"Laura",
""
],
[
"Pujol",
"Oriol",
""
],
[
"Moya",
"Josep",
""
],
[
"Violant",
"Verónica",
""
],
[
"Anguera",
"María Teresa",
""
]
] | TITLE: A Gesture Recognition System for Detecting Behavioral Patterns of ADHD
ABSTRACT: We present an application of gesture recognition using an extension of
Dynamic Time Warping (DTW) to recognize behavioural patterns of Attention
Deficit Hyperactivity Disorder (ADHD). We propose an extension of DTW using
one-class classifiers in order to be able to encode the variability of a
gesture category, and thus, perform an alignment between a gesture sample and a
gesture class. We model the set of gesture samples of a certain gesture
category using either GMMs or an approximation of Convex Hulls. Thus, we add a
theoretical contribution to classical warping path in DTW by including local
modeling of intra-class gesture variability. This methodology is applied in a
clinical context, detecting a group of ADHD behavioural patterns defined by
experts in psychology/psychiatry, to provide support to clinicians in the
diagnose procedure. The proposed methodology is tested on a novel multi-modal
dataset (RGB plus Depth) of ADHD children recordings with behavioural patterns.
We obtain satisfying results when compared to standard state-of-the-art
approaches in the DTW context.
|
1411.1171 | Rui Zeng | Rui Zeng, Jiasong Wu, Zhuhong Shao, Lotfi Senhadji, and Huazhong Shu | Multilinear Principal Component Analysis Network for Tensor Object
Classification | 4 pages, 3 figures | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The recently proposed principal component analysis network (PCANet) has been
proved high performance for visual content classification. In this letter, we
develop a tensorial extension of PCANet, namely, multilinear principal analysis
component network (MPCANet), for tensor object classification. Compared to
PCANet, the proposed MPCANet uses the spatial structure and the relationship
between each dimension of tensor objects much more efficiently. Experiments
were conducted on different visual content datasets including UCF sports action
video sequences database and UCF11 database. The experimental results have
revealed that the proposed MPCANet achieves higher classification accuracy than
PCANet for tensor object classification.
| [
{
"version": "v1",
"created": "Wed, 5 Nov 2014 07:27:08 GMT"
}
] | 2014-11-06T00:00:00 | [
[
"Zeng",
"Rui",
""
],
[
"Wu",
"Jiasong",
""
],
[
"Shao",
"Zhuhong",
""
],
[
"Senhadji",
"Lotfi",
""
],
[
"Shu",
"Huazhong",
""
]
] | TITLE: Multilinear Principal Component Analysis Network for Tensor Object
Classification
ABSTRACT: The recently proposed principal component analysis network (PCANet) has been
proved high performance for visual content classification. In this letter, we
develop a tensorial extension of PCANet, namely, multilinear principal analysis
component network (MPCANet), for tensor object classification. Compared to
PCANet, the proposed MPCANet uses the spatial structure and the relationship
between each dimension of tensor objects much more efficiently. Experiments
were conducted on different visual content datasets including UCF sports action
video sequences database and UCF11 database. The experimental results have
revealed that the proposed MPCANet achieves higher classification accuracy than
PCANet for tensor object classification.
|
1411.1372 | Nima Keivan | Nima Keivan and Gabe Sibley | Online SLAM with Any-time Self-calibration and Automatic Change
Detection | 8 pages, 6 figures | null | null | null | cs.CV cs.RO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A framework for online simultaneous localization, mapping and
self-calibration is presented which can detect and handle significant change in
the calibration parameters. Estimates are computed in constant-time by
factoring the problem and focusing on segments of the trajectory that are most
informative for the purposes of calibration. A novel technique is presented to
detect the probability that a significant change is present in the calibration
parameters. The system is then able to re-calibrate. Maximum likelihood
trajectory and map estimates are computed using an asynchronous and adaptive
optimization. The system requires no prior information and is able to
initialize without any special motions or routines, or in the case where
observability over calibration parameters is delayed. The system is
experimentally validated to calibrate camera intrinsic parameters for a
nonlinear camera model on a monocular dataset featuring a significant zoom
event partway through, and achieves high accuracy despite unknown initial
calibration parameters. Self-calibration and re-calibration parameters are
shown to closely match estimates computed using a calibration target. The
accuracy of the system is demonstrated with SLAM results that achieve sub-1%
distance-travel error even in the presence of significant re-calibration
events.
| [
{
"version": "v1",
"created": "Wed, 5 Nov 2014 19:39:41 GMT"
}
] | 2014-11-06T00:00:00 | [
[
"Keivan",
"Nima",
""
],
[
"Sibley",
"Gabe",
""
]
] | TITLE: Online SLAM with Any-time Self-calibration and Automatic Change
Detection
ABSTRACT: A framework for online simultaneous localization, mapping and
self-calibration is presented which can detect and handle significant change in
the calibration parameters. Estimates are computed in constant-time by
factoring the problem and focusing on segments of the trajectory that are most
informative for the purposes of calibration. A novel technique is presented to
detect the probability that a significant change is present in the calibration
parameters. The system is then able to re-calibrate. Maximum likelihood
trajectory and map estimates are computed using an asynchronous and adaptive
optimization. The system requires no prior information and is able to
initialize without any special motions or routines, or in the case where
observability over calibration parameters is delayed. The system is
experimentally validated to calibrate camera intrinsic parameters for a
nonlinear camera model on a monocular dataset featuring a significant zoom
event partway through, and achieves high accuracy despite unknown initial
calibration parameters. Self-calibration and re-calibration parameters are
shown to closely match estimates computed using a calibration target. The
accuracy of the system is demonstrated with SLAM results that achieve sub-1%
distance-travel error even in the presence of significant re-calibration
events.
|
1406.1134 | Piotr Doll\'ar | Woonhyun Nam, Piotr Doll\'ar, Joon Hee Han | Local Decorrelation For Improved Detection | To appear in Neural Information Processing Systems (NIPS), 2014 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Even with the advent of more sophisticated, data-hungry methods, boosted
decision trees remain extraordinarily successful for fast rigid object
detection, achieving top accuracy on numerous datasets. While effective, most
boosted detectors use decision trees with orthogonal (single feature) splits,
and the topology of the resulting decision boundary may not be well matched to
the natural topology of the data. Given highly correlated data, decision trees
with oblique (multiple feature) splits can be effective. Use of oblique splits,
however, comes at considerable computational expense. Inspired by recent work
on discriminative decorrelation of HOG features, we instead propose an
efficient feature transform that removes correlations in local neighborhoods.
The result is an overcomplete but locally decorrelated representation ideally
suited for use with orthogonal decision trees. In fact, orthogonal trees with
our locally decorrelated features outperform oblique trees trained over the
original features at a fraction of the computational cost. The overall
improvement in accuracy is dramatic: on the Caltech Pedestrian Dataset, we
reduce false positives nearly tenfold over the previous state-of-the-art.
| [
{
"version": "v1",
"created": "Wed, 4 Jun 2014 18:20:38 GMT"
},
{
"version": "v2",
"created": "Tue, 4 Nov 2014 02:50:18 GMT"
}
] | 2014-11-05T00:00:00 | [
[
"Nam",
"Woonhyun",
""
],
[
"Dollár",
"Piotr",
""
],
[
"Han",
"Joon Hee",
""
]
] | TITLE: Local Decorrelation For Improved Detection
ABSTRACT: Even with the advent of more sophisticated, data-hungry methods, boosted
decision trees remain extraordinarily successful for fast rigid object
detection, achieving top accuracy on numerous datasets. While effective, most
boosted detectors use decision trees with orthogonal (single feature) splits,
and the topology of the resulting decision boundary may not be well matched to
the natural topology of the data. Given highly correlated data, decision trees
with oblique (multiple feature) splits can be effective. Use of oblique splits,
however, comes at considerable computational expense. Inspired by recent work
on discriminative decorrelation of HOG features, we instead propose an
efficient feature transform that removes correlations in local neighborhoods.
The result is an overcomplete but locally decorrelated representation ideally
suited for use with orthogonal decision trees. In fact, orthogonal trees with
our locally decorrelated features outperform oblique trees trained over the
original features at a fraction of the computational cost. The overall
improvement in accuracy is dramatic: on the Caltech Pedestrian Dataset, we
reduce false positives nearly tenfold over the previous state-of-the-art.
|
1407.3399 | Xianjie Chen | Xianjie Chen, Alan Yuille | Articulated Pose Estimation by a Graphical Model with Image Dependent
Pairwise Relations | NIPS 2014 Camera Ready | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We present a method for estimating articulated human pose from a single
static image based on a graphical model with novel pairwise relations that make
adaptive use of local image measurements. More precisely, we specify a
graphical model for human pose which exploits the fact the local image
measurements can be used both to detect parts (or joints) and also to predict
the spatial relationships between them (Image Dependent Pairwise Relations).
These spatial relationships are represented by a mixture model. We use Deep
Convolutional Neural Networks (DCNNs) to learn conditional probabilities for
the presence of parts and their spatial relationships within image patches.
Hence our model combines the representational flexibility of graphical models
with the efficiency and statistical power of DCNNs. Our method significantly
outperforms the state of the art methods on the LSP and FLIC datasets and also
performs very well on the Buffy dataset without any training.
| [
{
"version": "v1",
"created": "Sat, 12 Jul 2014 17:04:21 GMT"
},
{
"version": "v2",
"created": "Tue, 4 Nov 2014 17:28:15 GMT"
}
] | 2014-11-05T00:00:00 | [
[
"Chen",
"Xianjie",
""
],
[
"Yuille",
"Alan",
""
]
] | TITLE: Articulated Pose Estimation by a Graphical Model with Image Dependent
Pairwise Relations
ABSTRACT: We present a method for estimating articulated human pose from a single
static image based on a graphical model with novel pairwise relations that make
adaptive use of local image measurements. More precisely, we specify a
graphical model for human pose which exploits the fact the local image
measurements can be used both to detect parts (or joints) and also to predict
the spatial relationships between them (Image Dependent Pairwise Relations).
These spatial relationships are represented by a mixture model. We use Deep
Convolutional Neural Networks (DCNNs) to learn conditional probabilities for
the presence of parts and their spatial relationships within image patches.
Hence our model combines the representational flexibility of graphical models
with the efficiency and statistical power of DCNNs. Our method significantly
outperforms the state of the art methods on the LSP and FLIC datasets and also
performs very well on the Buffy dataset without any training.
|
1409.6075 | Tyler Ward | Tyler Ward | The Information Theoretically Efficient Model (ITEM): A model for
computerized analysis of large datasets | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This document discusses the Information Theoretically Efficient Model (ITEM),
a computerized system to generate an information theoretically efficient
multinomial logistic regression from a general dataset. More specifically, this
model is designed to succeed even where the logit transform of the dependent
variable is not necessarily linear in the independent variables. This research
shows that for large datasets, the resulting models can be produced on modern
computers in a tractable amount of time. These models are also resistant to
overfitting, and as such they tend to produce interpretable models with only a
limited number of features, all of which are designed to be well behaved.
| [
{
"version": "v1",
"created": "Mon, 22 Sep 2014 03:39:23 GMT"
},
{
"version": "v2",
"created": "Mon, 6 Oct 2014 11:12:07 GMT"
},
{
"version": "v3",
"created": "Tue, 4 Nov 2014 05:41:04 GMT"
}
] | 2014-11-05T00:00:00 | [
[
"Ward",
"Tyler",
""
]
] | TITLE: The Information Theoretically Efficient Model (ITEM): A model for
computerized analysis of large datasets
ABSTRACT: This document discusses the Information Theoretically Efficient Model (ITEM),
a computerized system to generate an information theoretically efficient
multinomial logistic regression from a general dataset. More specifically, this
model is designed to succeed even where the logit transform of the dependent
variable is not necessarily linear in the independent variables. This research
shows that for large datasets, the resulting models can be produced on modern
computers in a tractable amount of time. These models are also resistant to
overfitting, and as such they tend to produce interpretable models with only a
limited number of features, all of which are designed to be well behaved.
|
1411.0722 | Yaneer Bar-Yam | Urbano Fran\c{c}a, Hiroki Sayama, Colin McSwiggen, Roozbeh Daneshvar
and Yaneer Bar-Yam | Visualizing the "Heartbeat" of a City with Tweets | 11 pages, 6 figures | null | null | New England Complex Systems Institute Report 2014-11-01 | physics.soc-ph cs.CY cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Describing the dynamics of a city is a crucial step to both understanding the
human activity in urban environments and to planning and designing cities
accordingly. Here we describe the collective dynamics of New York City and
surrounding areas as seen through the lens of Twitter usage. In particular, we
observe and quantify the patterns that emerge naturally from the hourly
activities in different areas of New York City, and discuss how they can be
used to understand the urban areas. Using a dataset that includes more than 6
million geolocated Twitter messages we construct a movie of the geographic
density of tweets. We observe the diurnal "heartbeat" of the NYC area. The
largest scale dynamics are the waking and sleeping cycle and commuting from
residential communities to office areas in Manhattan. Hourly dynamics reflect
the interplay of commuting, work and leisure, including whether people are
preoccupied with other activities or actively using Twitter. Differences
between weekday and weekend dynamics point to changes in when people wake and
sleep, and engage in social activities. We show that by measuring the average
distances to the heart of the city one can quantify the weekly differences and
the shift in behavior during weekends. We also identify locations and times of
high Twitter activity that occur because of specific activities. These include
early morning high levels of traffic as people arrive and wait at air
transportation hubs, and on Sunday at the Meadowlands Sports Complex and Statue
of Liberty. We analyze the role of particular individuals where they have large
impacts on overall Twitter activity. Our analysis points to the opportunity to
develop insight into both geographic social dynamics and attention through
social media analysis.
| [
{
"version": "v1",
"created": "Mon, 3 Nov 2014 22:27:23 GMT"
}
] | 2014-11-05T00:00:00 | [
[
"França",
"Urbano",
""
],
[
"Sayama",
"Hiroki",
""
],
[
"McSwiggen",
"Colin",
""
],
[
"Daneshvar",
"Roozbeh",
""
],
[
"Bar-Yam",
"Yaneer",
""
]
] | TITLE: Visualizing the "Heartbeat" of a City with Tweets
ABSTRACT: Describing the dynamics of a city is a crucial step to both understanding the
human activity in urban environments and to planning and designing cities
accordingly. Here we describe the collective dynamics of New York City and
surrounding areas as seen through the lens of Twitter usage. In particular, we
observe and quantify the patterns that emerge naturally from the hourly
activities in different areas of New York City, and discuss how they can be
used to understand the urban areas. Using a dataset that includes more than 6
million geolocated Twitter messages we construct a movie of the geographic
density of tweets. We observe the diurnal "heartbeat" of the NYC area. The
largest scale dynamics are the waking and sleeping cycle and commuting from
residential communities to office areas in Manhattan. Hourly dynamics reflect
the interplay of commuting, work and leisure, including whether people are
preoccupied with other activities or actively using Twitter. Differences
between weekday and weekend dynamics point to changes in when people wake and
sleep, and engage in social activities. We show that by measuring the average
distances to the heart of the city one can quantify the weekly differences and
the shift in behavior during weekends. We also identify locations and times of
high Twitter activity that occur because of specific activities. These include
early morning high levels of traffic as people arrive and wait at air
transportation hubs, and on Sunday at the Meadowlands Sports Complex and Statue
of Liberty. We analyze the role of particular individuals where they have large
impacts on overall Twitter activity. Our analysis points to the opportunity to
develop insight into both geographic social dynamics and attention through
social media analysis.
|
1411.0860 | Miao Xu | Miao Xu, Rong Jin, Zhi-Hua Zhou | CUR Algorithm for Partially Observed Matrices | null | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | CUR matrix decomposition computes the low rank approximation of a given
matrix by using the actual rows and columns of the matrix. It has been a very
useful tool for handling large matrices. One limitation with the existing
algorithms for CUR matrix decomposition is that they need an access to the {\it
full} matrix, a requirement that can be difficult to fulfill in many real world
applications. In this work, we alleviate this limitation by developing a CUR
decomposition algorithm for partially observed matrices. In particular, the
proposed algorithm computes the low rank approximation of the target matrix
based on (i) the randomly sampled rows and columns, and (ii) a subset of
observed entries that are randomly sampled from the matrix. Our analysis shows
the relative error bound, measured by spectral norm, for the proposed algorithm
when the target matrix is of full rank. We also show that only $O(n r\ln r)$
observed entries are needed by the proposed algorithm to perfectly recover a
rank $r$ matrix of size $n\times n$, which improves the sample complexity of
the existing algorithms for matrix completion. Empirical studies on both
synthetic and real-world datasets verify our theoretical claims and demonstrate
the effectiveness of the proposed algorithm.
| [
{
"version": "v1",
"created": "Tue, 4 Nov 2014 11:03:50 GMT"
}
] | 2014-11-05T00:00:00 | [
[
"Xu",
"Miao",
""
],
[
"Jin",
"Rong",
""
],
[
"Zhou",
"Zhi-Hua",
""
]
] | TITLE: CUR Algorithm for Partially Observed Matrices
ABSTRACT: CUR matrix decomposition computes the low rank approximation of a given
matrix by using the actual rows and columns of the matrix. It has been a very
useful tool for handling large matrices. One limitation with the existing
algorithms for CUR matrix decomposition is that they need an access to the {\it
full} matrix, a requirement that can be difficult to fulfill in many real world
applications. In this work, we alleviate this limitation by developing a CUR
decomposition algorithm for partially observed matrices. In particular, the
proposed algorithm computes the low rank approximation of the target matrix
based on (i) the randomly sampled rows and columns, and (ii) a subset of
observed entries that are randomly sampled from the matrix. Our analysis shows
the relative error bound, measured by spectral norm, for the proposed algorithm
when the target matrix is of full rank. We also show that only $O(n r\ln r)$
observed entries are needed by the proposed algorithm to perfectly recover a
rank $r$ matrix of size $n\times n$, which improves the sample complexity of
the existing algorithms for matrix completion. Empirical studies on both
synthetic and real-world datasets verify our theoretical claims and demonstrate
the effectiveness of the proposed algorithm.
|
1309.0326 | Micha{\l} {\L}opuszy\'nski | Micha{\l} {\L}opuszy\'nski, {\L}ukasz Bolikowski | Tagging Scientific Publications using Wikipedia and Natural Language
Processing Tools. Comparison on the ArXiv Dataset | null | Communications in Computer and Information Science Volume 416,
Springer 2014, pp 16-27 | 10.1007/978-3-319-08425-1_3 | null | cs.CL cs.DL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we compare two simple methods of tagging scientific
publications with labels reflecting their content. As a first source of labels
Wikipedia is employed, second label set is constructed from the noun phrases
occurring in the analyzed corpus. We examine the statistical properties and the
effectiveness of both approaches on the dataset consisting of abstracts from
0.7 million of scientific documents deposited in the ArXiv preprint collection.
We believe that obtained tags can be later on applied as useful document
features in various machine learning tasks (document similarity, clustering,
topic modelling, etc.).
| [
{
"version": "v1",
"created": "Mon, 2 Sep 2013 09:09:27 GMT"
},
{
"version": "v2",
"created": "Wed, 13 Aug 2014 14:30:21 GMT"
},
{
"version": "v3",
"created": "Mon, 3 Nov 2014 14:48:29 GMT"
}
] | 2014-11-04T00:00:00 | [
[
"Łopuszyński",
"Michał",
""
],
[
"Bolikowski",
"Łukasz",
""
]
] | TITLE: Tagging Scientific Publications using Wikipedia and Natural Language
Processing Tools. Comparison on the ArXiv Dataset
ABSTRACT: In this work, we compare two simple methods of tagging scientific
publications with labels reflecting their content. As a first source of labels
Wikipedia is employed, second label set is constructed from the noun phrases
occurring in the analyzed corpus. We examine the statistical properties and the
effectiveness of both approaches on the dataset consisting of abstracts from
0.7 million of scientific documents deposited in the ArXiv preprint collection.
We believe that obtained tags can be later on applied as useful document
features in various machine learning tasks (document similarity, clustering,
topic modelling, etc.).
|
1411.0052 | Chris Muelder | Arnaud Sallaberry, Yang-Chih Fu, Hwai-Chung Ho, Kwan-Liu Ma | ContactTrees: A Technique for Studying Personal Network Data | null | null | null | null | cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Network visualization allows a quick glance at how nodes (or actors) are
connected by edges (or ties). A conventional network diagram of "contact tree"
maps out a root and branches that represent the structure of nodes and edges,
often without further specifying leaves or fruits that would have grown from
small branches. By furnishing such a network structure with leaves and fruits,
we reveal details about "contacts" in our ContactTrees that underline ties and
relationships. Our elegant design employs a bottom-up approach that resembles a
recent attempt to understand subjective well-being by means of a series of
emotions. Such a bottom-up approach to social-network studies decomposes each
tie into a series of interactions or contacts, which help deepen our
understanding of the complexity embedded in a network structure. Unlike
previous network visualizations, ContactTrees can highlight how relationships
form and change based upon interactions among actors, and how relationships and
networks vary by contact attributes. Based on a botanical tree metaphor, the
design is easy to construct and the resulting tree-like visualization can
display many properties at both tie and contact levels, a key ingredient
missing from conventional techniques of network visualization. We first
demonstrate ContactTrees using a dataset consisting of three waves of 3-month
contact diaries over the 2004-2012 period, then compare ContactTrees with
alternative tools and discuss how this tool can be applied to other types of
datasets.
| [
{
"version": "v1",
"created": "Sat, 1 Nov 2014 01:44:15 GMT"
}
] | 2014-11-04T00:00:00 | [
[
"Sallaberry",
"Arnaud",
""
],
[
"Fu",
"Yang-Chih",
""
],
[
"Ho",
"Hwai-Chung",
""
],
[
"Ma",
"Kwan-Liu",
""
]
] | TITLE: ContactTrees: A Technique for Studying Personal Network Data
ABSTRACT: Network visualization allows a quick glance at how nodes (or actors) are
connected by edges (or ties). A conventional network diagram of "contact tree"
maps out a root and branches that represent the structure of nodes and edges,
often without further specifying leaves or fruits that would have grown from
small branches. By furnishing such a network structure with leaves and fruits,
we reveal details about "contacts" in our ContactTrees that underline ties and
relationships. Our elegant design employs a bottom-up approach that resembles a
recent attempt to understand subjective well-being by means of a series of
emotions. Such a bottom-up approach to social-network studies decomposes each
tie into a series of interactions or contacts, which help deepen our
understanding of the complexity embedded in a network structure. Unlike
previous network visualizations, ContactTrees can highlight how relationships
form and change based upon interactions among actors, and how relationships and
networks vary by contact attributes. Based on a botanical tree metaphor, the
design is easy to construct and the resulting tree-like visualization can
display many properties at both tie and contact levels, a key ingredient
missing from conventional techniques of network visualization. We first
demonstrate ContactTrees using a dataset consisting of three waves of 3-month
contact diaries over the 2004-2012 period, then compare ContactTrees with
alternative tools and discuss how this tool can be applied to other types of
datasets.
|
1411.0126 | GowthamRangarajan Raman | Gowtham Rangarajan Raman | Detection of texts in natural images | null | null | null | null | cs.CV | http://creativecommons.org/licenses/by/3.0/ | A framework that makes use of Connected components and supervised Support
machine to recognise texts is proposed. The image is preprocessed and and edge
graph is calculated using a probabilistic framework to compensate for
photometric noise. Connected components over the resultant image is calculated,
which is bounded and then pruned using geometric constraints. Finally a Gabor
Feature based SVM is used to classify the presence of text in the candidates.
The proposed method was tested with ICDAR 10 dataset and few other images
available on the internet. It resulted in a recall and precision metric of 0.72
and 0.88 comfortably better than the benchmark Eiphstein's algorithm. The
proposed method recorded a 0.70 and 0.74 in natural images which is
significantly better than current methods on natural images. The proposed
method also scales almost linearly for high resolution, cluttered images.
| [
{
"version": "v1",
"created": "Sat, 1 Nov 2014 15:06:23 GMT"
}
] | 2014-11-04T00:00:00 | [
[
"Raman",
"Gowtham Rangarajan",
""
]
] | TITLE: Detection of texts in natural images
ABSTRACT: A framework that makes use of Connected components and supervised Support
machine to recognise texts is proposed. The image is preprocessed and and edge
graph is calculated using a probabilistic framework to compensate for
photometric noise. Connected components over the resultant image is calculated,
which is bounded and then pruned using geometric constraints. Finally a Gabor
Feature based SVM is used to classify the presence of text in the candidates.
The proposed method was tested with ICDAR 10 dataset and few other images
available on the internet. It resulted in a recall and precision metric of 0.72
and 0.88 comfortably better than the benchmark Eiphstein's algorithm. The
proposed method recorded a 0.70 and 0.74 in natural images which is
significantly better than current methods on natural images. The proposed
method also scales almost linearly for high resolution, cluttered images.
|
1411.0392 | Roozbeh Rajabi | Roozbeh Rajabi, Hassan Ghassemian | Sparsity Constrained Graph Regularized NMF for Spectral Unmixing of
Hyperspectral Data | 10 pages, Journal | null | 10.1007/s12524-014-0408-2 | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Hyperspectral images contain mixed pixels due to low spatial resolution of
hyperspectral sensors. Mixed pixels are pixels containing more than one
distinct material called endmembers. The presence percentages of endmembers in
mixed pixels are called abundance fractions. Spectral unmixing problem refers
to decomposing these pixels into a set of endmembers and abundance fractions.
Due to nonnegativity constraint on abundance fractions, nonnegative matrix
factorization methods (NMF) have been widely used for solving spectral unmixing
problem. In this paper we have used graph regularized NMF (GNMF) method
combined with sparseness constraint to decompose mixed pixels in hyperspectral
imagery. This method preserves the geometrical structure of data while
representing it in low dimensional space. Adaptive regularization parameter
based on temperature schedule in simulated annealing method also has been used
in this paper for the sparseness term. Proposed algorithm is applied on
synthetic and real datasets. Synthetic data is generated based on endmembers
from USGS spectral library. AVIRIS Cuprite dataset is used as real dataset for
evaluation of proposed method. Results are quantified based on spectral angle
distance (SAD) and abundance angle distance (AAD) measures. Results in
comparison with other methods show that the proposed method can unmix data more
effectively. Specifically for the Cuprite dataset, performance of the proposed
method is approximately 10% better than the VCA and Sparse NMF in terms of root
mean square of SAD.
| [
{
"version": "v1",
"created": "Mon, 3 Nov 2014 08:41:32 GMT"
}
] | 2014-11-04T00:00:00 | [
[
"Rajabi",
"Roozbeh",
""
],
[
"Ghassemian",
"Hassan",
""
]
] | TITLE: Sparsity Constrained Graph Regularized NMF for Spectral Unmixing of
Hyperspectral Data
ABSTRACT: Hyperspectral images contain mixed pixels due to low spatial resolution of
hyperspectral sensors. Mixed pixels are pixels containing more than one
distinct material called endmembers. The presence percentages of endmembers in
mixed pixels are called abundance fractions. Spectral unmixing problem refers
to decomposing these pixels into a set of endmembers and abundance fractions.
Due to nonnegativity constraint on abundance fractions, nonnegative matrix
factorization methods (NMF) have been widely used for solving spectral unmixing
problem. In this paper we have used graph regularized NMF (GNMF) method
combined with sparseness constraint to decompose mixed pixels in hyperspectral
imagery. This method preserves the geometrical structure of data while
representing it in low dimensional space. Adaptive regularization parameter
based on temperature schedule in simulated annealing method also has been used
in this paper for the sparseness term. Proposed algorithm is applied on
synthetic and real datasets. Synthetic data is generated based on endmembers
from USGS spectral library. AVIRIS Cuprite dataset is used as real dataset for
evaluation of proposed method. Results are quantified based on spectral angle
distance (SAD) and abundance angle distance (AAD) measures. Results in
comparison with other methods show that the proposed method can unmix data more
effectively. Specifically for the Cuprite dataset, performance of the proposed
method is approximately 10% better than the VCA and Sparse NMF in terms of root
mean square of SAD.
|
1411.0591 | Charles Fisher | Charles K. Fisher and Pankaj Mehta | Bayesian feature selection with strongly-regularizing priors maps to the
Ising Model | null | null | null | null | cond-mat.stat-mech cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Identifying small subsets of features that are relevant for prediction and/or
classification tasks is a central problem in machine learning and statistics.
The feature selection task is especially important, and computationally
difficult, for modern datasets where the number of features can be comparable
to, or even exceed, the number of samples. Here, we show that feature selection
with Bayesian inference takes a universal form and reduces to calculating the
magnetizations of an Ising model, under some mild conditions. Our results
exploit the observation that the evidence takes a universal form for
strongly-regularizing priors --- priors that have a large effect on the
posterior probability even in the infinite data limit. We derive explicit
expressions for feature selection for generalized linear models, a large class
of statistical techniques that include linear and logistic regression. We
illustrate the power of our approach by analyzing feature selection in a
logistic regression-based classifier trained to distinguish between the letters
B and D in the notMNIST dataset.
| [
{
"version": "v1",
"created": "Mon, 3 Nov 2014 18:15:29 GMT"
}
] | 2014-11-04T00:00:00 | [
[
"Fisher",
"Charles K.",
""
],
[
"Mehta",
"Pankaj",
""
]
] | TITLE: Bayesian feature selection with strongly-regularizing priors maps to the
Ising Model
ABSTRACT: Identifying small subsets of features that are relevant for prediction and/or
classification tasks is a central problem in machine learning and statistics.
The feature selection task is especially important, and computationally
difficult, for modern datasets where the number of features can be comparable
to, or even exceed, the number of samples. Here, we show that feature selection
with Bayesian inference takes a universal form and reduces to calculating the
magnetizations of an Ising model, under some mild conditions. Our results
exploit the observation that the evidence takes a universal form for
strongly-regularizing priors --- priors that have a large effect on the
posterior probability even in the infinite data limit. We derive explicit
expressions for feature selection for generalized linear models, a large class
of statistical techniques that include linear and logistic regression. We
illustrate the power of our approach by analyzing feature selection in a
logistic regression-based classifier trained to distinguish between the letters
B and D in the notMNIST dataset.
|
1310.7616 | Jinsub Kim | Jinsub Kim, Lang Tong, Robert J. Thomas | Data Framing Attack on State Estimation | To appear in IEEE Journal on Selected Areas in Communications:
Special series on Smart Grid Communications, 2014 | IEEE Journal on Selected Areas in Communications, vol. 32, no. 7,
pp 1460-1470, 2014 | 10.1109/JSAC.2014.2332032 | null | cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | A new mechanism aimed at misleading a power system control center about the
source of a data attack is proposed. As a man-in-the-middle state attack, a
data framing attack is proposed to exploit the bad data detection and
identification mechanisms currently in use at most control centers. In
particular, the proposed attack frames meters that are providing correct data
as sources of bad data such that the control center will remove useful
measurements that would otherwise be used by the state estimator.
The optimal design of a data framing attack is formulated as a quadratically
constrained quadratic program (QCQP). It is shown that the proposed attack is
capable of perturbing the power system state estimate by an arbitrary degree
controlling only half of a critical set of measurements that are needed to make
a system unobservable. Implications of this attack on power system operations
are discussed, and the attack performance is evaluated using benchmark systems.
| [
{
"version": "v1",
"created": "Mon, 28 Oct 2013 20:41:16 GMT"
},
{
"version": "v2",
"created": "Mon, 21 Apr 2014 04:07:59 GMT"
}
] | 2014-11-03T00:00:00 | [
[
"Kim",
"Jinsub",
""
],
[
"Tong",
"Lang",
""
],
[
"Thomas",
"Robert J.",
""
]
] | TITLE: Data Framing Attack on State Estimation
ABSTRACT: A new mechanism aimed at misleading a power system control center about the
source of a data attack is proposed. As a man-in-the-middle state attack, a
data framing attack is proposed to exploit the bad data detection and
identification mechanisms currently in use at most control centers. In
particular, the proposed attack frames meters that are providing correct data
as sources of bad data such that the control center will remove useful
measurements that would otherwise be used by the state estimator.
The optimal design of a data framing attack is formulated as a quadratically
constrained quadratic program (QCQP). It is shown that the proposed attack is
capable of perturbing the power system state estimate by an arbitrary degree
controlling only half of a critical set of measurements that are needed to make
a system unobservable. Implications of this attack on power system operations
are discussed, and the attack performance is evaluated using benchmark systems.
|
1410.6858 | Alberto Dainotti | Alberto Dainotti, Karyn Benson, Alistair King, kc claffy, Eduard
Glatz, Xenofontas Dimitropoulos, Philipp Richter, Alessandro Finamore, Alex
C. Snoeren | Lost in Space: Improving Inference of IPv4 Address Space Utilization | null | null | null | null | cs.NI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | One challenge in understanding the evolution of Internet infrastructure is
the lack of systematic mechanisms for monitoring the extent to which allocated
IP addresses are actually used. In this paper we try to advance the science of
inferring IPv4 address space utilization by analyzing and correlating results
obtained through different types of measurements. We have previously studied an
approach based on passive measurements that can reveal used portions of the
address space unseen by active approaches. In this paper, we study such passive
approaches in detail, extending our methodology to four different types of
vantage points, identifying traffic components that most significantly
contribute to discovering used IPv4 network blocks. We then combine the results
we obtained through passive measurements together with data from active
measurement studies, as well as measurements from BGP and additional datasets
available to researchers. Through the analysis of this large collection of
heterogeneous datasets, we substantially improve the state of the art in terms
of: (i) understanding the challenges and opportunities in using passive and
active techniques to study address utilization; and (ii) knowledge of the
utilization of the IPv4 space.
| [
{
"version": "v1",
"created": "Sat, 25 Oct 2014 00:29:54 GMT"
},
{
"version": "v2",
"created": "Thu, 30 Oct 2014 22:07:44 GMT"
}
] | 2014-11-03T00:00:00 | [
[
"Dainotti",
"Alberto",
""
],
[
"Benson",
"Karyn",
""
],
[
"King",
"Alistair",
""
],
[
"claffy",
"kc",
""
],
[
"Glatz",
"Eduard",
""
],
[
"Dimitropoulos",
"Xenofontas",
""
],
[
"Richter",
"Philipp",
""
],
[
"Finamore",
"Alessandro",
""
],
[
"Snoeren",
"Alex C.",
""
]
] | TITLE: Lost in Space: Improving Inference of IPv4 Address Space Utilization
ABSTRACT: One challenge in understanding the evolution of Internet infrastructure is
the lack of systematic mechanisms for monitoring the extent to which allocated
IP addresses are actually used. In this paper we try to advance the science of
inferring IPv4 address space utilization by analyzing and correlating results
obtained through different types of measurements. We have previously studied an
approach based on passive measurements that can reveal used portions of the
address space unseen by active approaches. In this paper, we study such passive
approaches in detail, extending our methodology to four different types of
vantage points, identifying traffic components that most significantly
contribute to discovering used IPv4 network blocks. We then combine the results
we obtained through passive measurements together with data from active
measurement studies, as well as measurements from BGP and additional datasets
available to researchers. Through the analysis of this large collection of
heterogeneous datasets, we substantially improve the state of the art in terms
of: (i) understanding the challenges and opportunities in using passive and
active techniques to study address utilization; and (ii) knowledge of the
utilization of the IPv4 space.
|
1410.8586 | Tao Chen | Tao Chen, Damian Borth, Trevor Darrell and Shih-Fu Chang | DeepSentiBank: Visual Sentiment Concept Classification with Deep
Convolutional Neural Networks | 7 pages, 4 figures | null | null | null | cs.CV cs.LG cs.MM cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | This paper introduces a visual sentiment concept classification method based
on deep convolutional neural networks (CNNs). The visual sentiment concepts are
adjective noun pairs (ANPs) automatically discovered from the tags of web
photos, and can be utilized as effective statistical cues for detecting
emotions depicted in the images. Nearly one million Flickr images tagged with
these ANPs are downloaded to train the classifiers of the concepts. We adopt
the popular model of deep convolutional neural networks which recently shows
great performance improvement on classifying large-scale web-based image
dataset such as ImageNet. Our deep CNNs model is trained based on Caffe, a
newly developed deep learning framework. To deal with the biased training data
which only contains images with strong sentiment and to prevent overfitting, we
initialize the model with the model weights trained from ImageNet. Performance
evaluation shows the newly trained deep CNNs model SentiBank 2.0 (or called
DeepSentiBank) is significantly improved in both annotation accuracy and
retrieval performance, compared to its predecessors which mainly use binary SVM
classification models.
| [
{
"version": "v1",
"created": "Thu, 30 Oct 2014 22:57:12 GMT"
}
] | 2014-11-03T00:00:00 | [
[
"Chen",
"Tao",
""
],
[
"Borth",
"Damian",
""
],
[
"Darrell",
"Trevor",
""
],
[
"Chang",
"Shih-Fu",
""
]
] | TITLE: DeepSentiBank: Visual Sentiment Concept Classification with Deep
Convolutional Neural Networks
ABSTRACT: This paper introduces a visual sentiment concept classification method based
on deep convolutional neural networks (CNNs). The visual sentiment concepts are
adjective noun pairs (ANPs) automatically discovered from the tags of web
photos, and can be utilized as effective statistical cues for detecting
emotions depicted in the images. Nearly one million Flickr images tagged with
these ANPs are downloaded to train the classifiers of the concepts. We adopt
the popular model of deep convolutional neural networks which recently shows
great performance improvement on classifying large-scale web-based image
dataset such as ImageNet. Our deep CNNs model is trained based on Caffe, a
newly developed deep learning framework. To deal with the biased training data
which only contains images with strong sentiment and to prevent overfitting, we
initialize the model with the model weights trained from ImageNet. Performance
evaluation shows the newly trained deep CNNs model SentiBank 2.0 (or called
DeepSentiBank) is significantly improved in both annotation accuracy and
retrieval performance, compared to its predecessors which mainly use binary SVM
classification models.
|
1410.8664 | Yishi Lin | Yishi Lin, John C.S. Lui | Algorithmic Design for Competitive Influence Maximization Problems | null | null | null | null | cs.SI cs.DS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Given the popularity of the viral marketing campaign in online social
networks, finding an effective method to identify a set of most influential
nodes so to compete well with other viral marketing competitors is of upmost
importance. We propose a "General Competitive Independent Cascade (GCIC)" model
to describe the general influence propagation of two competing sources in the
same network. We formulate the "Competitive Influence Maximization (CIM)"
problem as follows: Under a prespecified influence propagation model and that
the competitor's seed set is known, how to find a seed set of $k$ nodes so as
to trigger the largest influence cascade? We propose a general algorithmic
framework TCIM for the CIM problem under the GCIC model. TCIM returns a
$(1-1/e-\epsilon)$-approximate solution with probability at least
$1-n^{-\ell}$, and has an efficient time complexity of $O(c(k+\ell)(m+n)\log
n/\epsilon^2)$, where $c$ depends on specific propagation model and may also
depend on $k$ and underlying network $G$. To the best of our knowledge, this is
the first general algorithmic framework that has both $(1-1/e-\epsilon)$
performance guarantee and practical efficiency. We conduct extensive
experiments on real-world datasets under three specific influence propagation
models, and show the efficiency and accuracy of our framework. In particular,
we achieve up to four orders of magnitude speedup as compared to the previous
state-of-the-art algorithms with the approximate guarantee.
| [
{
"version": "v1",
"created": "Fri, 31 Oct 2014 08:16:20 GMT"
}
] | 2014-11-03T00:00:00 | [
[
"Lin",
"Yishi",
""
],
[
"Lui",
"John C. S.",
""
]
] | TITLE: Algorithmic Design for Competitive Influence Maximization Problems
ABSTRACT: Given the popularity of the viral marketing campaign in online social
networks, finding an effective method to identify a set of most influential
nodes so to compete well with other viral marketing competitors is of upmost
importance. We propose a "General Competitive Independent Cascade (GCIC)" model
to describe the general influence propagation of two competing sources in the
same network. We formulate the "Competitive Influence Maximization (CIM)"
problem as follows: Under a prespecified influence propagation model and that
the competitor's seed set is known, how to find a seed set of $k$ nodes so as
to trigger the largest influence cascade? We propose a general algorithmic
framework TCIM for the CIM problem under the GCIC model. TCIM returns a
$(1-1/e-\epsilon)$-approximate solution with probability at least
$1-n^{-\ell}$, and has an efficient time complexity of $O(c(k+\ell)(m+n)\log
n/\epsilon^2)$, where $c$ depends on specific propagation model and may also
depend on $k$ and underlying network $G$. To the best of our knowledge, this is
the first general algorithmic framework that has both $(1-1/e-\epsilon)$
performance guarantee and practical efficiency. We conduct extensive
experiments on real-world datasets under three specific influence propagation
models, and show the efficiency and accuracy of our framework. In particular,
we achieve up to four orders of magnitude speedup as compared to the previous
state-of-the-art algorithms with the approximate guarantee.
|
1401.0733 | Ahmet Iscen | Ahmet Iscen, Eren Golge, Ilker Sarac, Pinar Duygulu | ConceptVision: A Flexible Scene Classification Framework | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We introduce ConceptVision, a method that aims for high accuracy in
categorizing large number of scenes, while keeping the model relatively simpler
and efficient for scalability. The proposed method combines the advantages of
both low-level representations and high-level semantic categories, and
eliminates the distinctions between different levels through the definition of
concepts. The proposed framework encodes the perspectives brought through
different concepts by considering them in concept groups. Different
perspectives are ensembled for the final decision. Extensive experiments are
carried out on benchmark datasets to test the effects of different concepts,
and methods used to ensemble. Comparisons with state-of-the-art studies show
that we can achieve better results with incorporation of concepts in different
levels with different perspectives.
| [
{
"version": "v1",
"created": "Fri, 3 Jan 2014 21:15:13 GMT"
},
{
"version": "v2",
"created": "Wed, 29 Oct 2014 20:19:35 GMT"
}
] | 2014-10-31T00:00:00 | [
[
"Iscen",
"Ahmet",
""
],
[
"Golge",
"Eren",
""
],
[
"Sarac",
"Ilker",
""
],
[
"Duygulu",
"Pinar",
""
]
] | TITLE: ConceptVision: A Flexible Scene Classification Framework
ABSTRACT: We introduce ConceptVision, a method that aims for high accuracy in
categorizing large number of scenes, while keeping the model relatively simpler
and efficient for scalability. The proposed method combines the advantages of
both low-level representations and high-level semantic categories, and
eliminates the distinctions between different levels through the definition of
concepts. The proposed framework encodes the perspectives brought through
different concepts by considering them in concept groups. Different
perspectives are ensembled for the final decision. Extensive experiments are
carried out on benchmark datasets to test the effects of different concepts,
and methods used to ensemble. Comparisons with state-of-the-art studies show
that we can achieve better results with incorporation of concepts in different
levels with different perspectives.
|
1407.7644 | Ariel Jaffe | Ariel Jaffe, Boaz Nadler and Yuval Kluger | Estimating the Accuracies of Multiple Classifiers Without Labeled Data | null | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In various situations one is given only the predictions of multiple
classifiers over a large unlabeled test data. This scenario raises the
following questions: Without any labeled data and without any a-priori
knowledge about the reliability of these different classifiers, is it possible
to consistently and computationally efficiently estimate their accuracies?
Furthermore, also in a completely unsupervised manner, can one construct a more
accurate unsupervised ensemble classifier? In this paper, focusing on the
binary case, we present simple, computationally efficient algorithms to solve
these questions. Furthermore, under standard classifier independence
assumptions, we prove our methods are consistent and study their asymptotic
error. Our approach is spectral, based on the fact that the off-diagonal
entries of the classifiers' covariance matrix and 3-d tensor are rank-one. We
illustrate the competitive performance of our algorithms via extensive
experiments on both artificial and real datasets.
| [
{
"version": "v1",
"created": "Tue, 29 Jul 2014 07:19:08 GMT"
},
{
"version": "v2",
"created": "Thu, 30 Oct 2014 11:23:37 GMT"
}
] | 2014-10-31T00:00:00 | [
[
"Jaffe",
"Ariel",
""
],
[
"Nadler",
"Boaz",
""
],
[
"Kluger",
"Yuval",
""
]
] | TITLE: Estimating the Accuracies of Multiple Classifiers Without Labeled Data
ABSTRACT: In various situations one is given only the predictions of multiple
classifiers over a large unlabeled test data. This scenario raises the
following questions: Without any labeled data and without any a-priori
knowledge about the reliability of these different classifiers, is it possible
to consistently and computationally efficiently estimate their accuracies?
Furthermore, also in a completely unsupervised manner, can one construct a more
accurate unsupervised ensemble classifier? In this paper, focusing on the
binary case, we present simple, computationally efficient algorithms to solve
these questions. Furthermore, under standard classifier independence
assumptions, we prove our methods are consistent and study their asymptotic
error. Our approach is spectral, based on the fact that the off-diagonal
entries of the classifiers' covariance matrix and 3-d tensor are rank-one. We
illustrate the competitive performance of our algorithms via extensive
experiments on both artificial and real datasets.
|
1410.8507 | Mark Taylor | M. B. Taylor | External Use of TOPCAT's Plotting Library | 4 pages, 1 figure | null | null | null | astro-ph.IM cs.MS | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The table analysis application TOPCAT uses a custom Java plotting library for
highly configurable high-performance interactive or exported visualisations in
two and three dimensions. We present here a variety of ways for end users or
application developers to make use of this library outside of the TOPCAT
application: via the command-line suite STILTS or its Jython variant JyStilts,
via a traditional Java API, or by programmatically assigning values to a set of
parameters in java code or using some form of inter-process communication. The
library has been built with large datasets in mind; interactive plots scale
well up to several million points, and static output to standard graphics
formats is possible for unlimited sized input data.
| [
{
"version": "v1",
"created": "Thu, 30 Oct 2014 19:29:08 GMT"
}
] | 2014-10-31T00:00:00 | [
[
"Taylor",
"M. B.",
""
]
] | TITLE: External Use of TOPCAT's Plotting Library
ABSTRACT: The table analysis application TOPCAT uses a custom Java plotting library for
highly configurable high-performance interactive or exported visualisations in
two and three dimensions. We present here a variety of ways for end users or
application developers to make use of this library outside of the TOPCAT
application: via the command-line suite STILTS or its Jython variant JyStilts,
via a traditional Java API, or by programmatically assigning values to a set of
parameters in java code or using some form of inter-process communication. The
library has been built with large datasets in mind; interactive plots scale
well up to several million points, and static output to standard graphics
formats is possible for unlimited sized input data.
|
1410.7709 | Tuomo Sipola | Antti Juvonen and Tuomo Sipola | Anomaly Detection Framework Using Rule Extraction for Efficient
Intrusion Detection | 35 pages, 12 figures, 7 tables | null | null | null | cs.LG cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Huge datasets in cyber security, such as network traffic logs, can be
analyzed using machine learning and data mining methods. However, the amount of
collected data is increasing, which makes analysis more difficult. Many machine
learning methods have not been designed for big datasets, and consequently are
slow and difficult to understand. We address the issue of efficient network
traffic classification by creating an intrusion detection framework that
applies dimensionality reduction and conjunctive rule extraction. The system
can perform unsupervised anomaly detection and use this information to create
conjunctive rules that classify huge amounts of traffic in real time. We test
the implemented system with the widely used KDD Cup 99 dataset and real-world
network logs to confirm that the performance is satisfactory. This system is
transparent and does not work like a black box, making it intuitive for domain
experts, such as network administrators.
| [
{
"version": "v1",
"created": "Tue, 28 Oct 2014 17:29:42 GMT"
}
] | 2014-10-30T00:00:00 | [
[
"Juvonen",
"Antti",
""
],
[
"Sipola",
"Tuomo",
""
]
] | TITLE: Anomaly Detection Framework Using Rule Extraction for Efficient
Intrusion Detection
ABSTRACT: Huge datasets in cyber security, such as network traffic logs, can be
analyzed using machine learning and data mining methods. However, the amount of
collected data is increasing, which makes analysis more difficult. Many machine
learning methods have not been designed for big datasets, and consequently are
slow and difficult to understand. We address the issue of efficient network
traffic classification by creating an intrusion detection framework that
applies dimensionality reduction and conjunctive rule extraction. The system
can perform unsupervised anomaly detection and use this information to create
conjunctive rules that classify huge amounts of traffic in real time. We test
the implemented system with the widely used KDD Cup 99 dataset and real-world
network logs to confirm that the performance is satisfactory. This system is
transparent and does not work like a black box, making it intuitive for domain
experts, such as network administrators.
|
1410.8034 | Xudong Liu | Xudong Liu, Bin Zhang, Ting Zhang and Chang Liu | Latent Feature Based FM Model For Rating Prediction | 4 pages, 3 figures, Large Scale Recommender Systems:workshop of
Recsys 2014 | null | null | null | cs.LG cs.IR stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Rating Prediction is a basic problem in Recommender System, and one of the
most widely used method is Factorization Machines(FM). However, traditional
matrix factorization methods fail to utilize the benefit of implicit feedback,
which has been proved to be important in Rating Prediction problem. In this
work, we consider a specific situation, movie rating prediction, where we
assume that watching history has a big influence on his/her rating behavior on
an item. We introduce two models, Latent Dirichlet Allocation(LDA) and
word2vec, both of which perform state-of-the-art results in training latent
features. Based on that, we propose two feature based models. One is the
Topic-based FM Model which provides the implicit feedback to the matrix
factorization. The other is the Vector-based FM Model which expresses the order
info of watching history. Empirical results on three datasets demonstrate that
our method performs better than the baseline model and confirm that
Vector-based FM Model usually works better as it contains the order info.
| [
{
"version": "v1",
"created": "Wed, 29 Oct 2014 15:51:54 GMT"
}
] | 2014-10-30T00:00:00 | [
[
"Liu",
"Xudong",
""
],
[
"Zhang",
"Bin",
""
],
[
"Zhang",
"Ting",
""
],
[
"Liu",
"Chang",
""
]
] | TITLE: Latent Feature Based FM Model For Rating Prediction
ABSTRACT: Rating Prediction is a basic problem in Recommender System, and one of the
most widely used method is Factorization Machines(FM). However, traditional
matrix factorization methods fail to utilize the benefit of implicit feedback,
which has been proved to be important in Rating Prediction problem. In this
work, we consider a specific situation, movie rating prediction, where we
assume that watching history has a big influence on his/her rating behavior on
an item. We introduce two models, Latent Dirichlet Allocation(LDA) and
word2vec, both of which perform state-of-the-art results in training latent
features. Based on that, we propose two feature based models. One is the
Topic-based FM Model which provides the implicit feedback to the matrix
factorization. The other is the Vector-based FM Model which expresses the order
info of watching history. Empirical results on three datasets demonstrate that
our method performs better than the baseline model and confirm that
Vector-based FM Model usually works better as it contains the order info.
|
1404.0545 | EPTCS | Thomas Given-Wilson (INRIA, Saclay) | An Intensional Concurrent Faithful Encoding of Turing Machines | In Proceedings ICE 2014, arXiv:1410.7013 | EPTCS 166, 2014, pp. 21-37 | 10.4204/EPTCS.166.4 | null | cs.FL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The benchmark for computation is typically given as Turing computability; the
ability for a computation to be performed by a Turing Machine. Many languages
exploit (indirect) encodings of Turing Machines to demonstrate their ability to
support arbitrary computation. However, these encodings are usually by
simulating the entire Turing Machine within the language, or by encoding a
language that does an encoding or simulation itself. This second category is
typical for process calculi that show an encoding of lambda-calculus (often
with restrictions) that in turn simulates a Turing Machine. Such approaches
lead to indirect encodings of Turing Machines that are complex, unclear, and
only weakly equivalent after computation. This paper presents an approach to
encoding Turing Machines into intensional process calculi that is faithful,
reduction preserving, and structurally equivalent. The encoding is demonstrated
in a simple asymmetric concurrent pattern calculus before generalised to
simplify infinite terms, and to show encodings into Concurrent Pattern Calculus
and Psi Calculi.
| [
{
"version": "v1",
"created": "Wed, 2 Apr 2014 13:25:43 GMT"
},
{
"version": "v2",
"created": "Wed, 7 May 2014 10:04:09 GMT"
},
{
"version": "v3",
"created": "Mon, 23 Jun 2014 09:06:05 GMT"
},
{
"version": "v4",
"created": "Tue, 28 Oct 2014 10:39:18 GMT"
}
] | 2014-10-29T00:00:00 | [
[
"Given-Wilson",
"Thomas",
"",
"INRIA, Saclay"
]
] | TITLE: An Intensional Concurrent Faithful Encoding of Turing Machines
ABSTRACT: The benchmark for computation is typically given as Turing computability; the
ability for a computation to be performed by a Turing Machine. Many languages
exploit (indirect) encodings of Turing Machines to demonstrate their ability to
support arbitrary computation. However, these encodings are usually by
simulating the entire Turing Machine within the language, or by encoding a
language that does an encoding or simulation itself. This second category is
typical for process calculi that show an encoding of lambda-calculus (often
with restrictions) that in turn simulates a Turing Machine. Such approaches
lead to indirect encodings of Turing Machines that are complex, unclear, and
only weakly equivalent after computation. This paper presents an approach to
encoding Turing Machines into intensional process calculi that is faithful,
reduction preserving, and structurally equivalent. The encoding is demonstrated
in a simple asymmetric concurrent pattern calculus before generalised to
simplify infinite terms, and to show encodings into Concurrent Pattern Calculus
and Psi Calculi.
|
1410.7414 | Junier Oliva | Junier Oliva, Willie Neiswanger, Barnabas Poczos, Eric Xing, Jeff
Schneider | Fast Function to Function Regression | null | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We analyze the problem of regression when both input covariates and output
responses are functions from a nonparametric function class. Function to
function regression (FFR) covers a large range of interesting applications
including time-series prediction problems, and also more general tasks like
studying a mapping between two separate types of distributions. However,
previous nonparametric estimators for FFR type problems scale badly
computationally with the number of input/output pairs in a data-set. Given the
complexity of a mapping between general functions it may be necessary to
consider large data-sets in order to achieve a low estimation risk. To address
this issue, we develop a novel scalable nonparametric estimator, the
Triple-Basis Estimator (3BE), which is capable of operating over datasets with
many instances. To the best of our knowledge, the 3BE is the first
nonparametric FFR estimator that can scale to massive datasets. We analyze the
3BE's risk and derive an upperbound rate. Furthermore, we show an improvement
of several orders of magnitude in terms of prediction speed and a reduction in
error over previous estimators in various real-world data-sets.
| [
{
"version": "v1",
"created": "Mon, 27 Oct 2014 20:15:18 GMT"
}
] | 2014-10-29T00:00:00 | [
[
"Oliva",
"Junier",
""
],
[
"Neiswanger",
"Willie",
""
],
[
"Poczos",
"Barnabas",
""
],
[
"Xing",
"Eric",
""
],
[
"Schneider",
"Jeff",
""
]
] | TITLE: Fast Function to Function Regression
ABSTRACT: We analyze the problem of regression when both input covariates and output
responses are functions from a nonparametric function class. Function to
function regression (FFR) covers a large range of interesting applications
including time-series prediction problems, and also more general tasks like
studying a mapping between two separate types of distributions. However,
previous nonparametric estimators for FFR type problems scale badly
computationally with the number of input/output pairs in a data-set. Given the
complexity of a mapping between general functions it may be necessary to
consider large data-sets in order to achieve a low estimation risk. To address
this issue, we develop a novel scalable nonparametric estimator, the
Triple-Basis Estimator (3BE), which is capable of operating over datasets with
many instances. To the best of our knowledge, the 3BE is the first
nonparametric FFR estimator that can scale to massive datasets. We analyze the
3BE's risk and derive an upperbound rate. Furthermore, we show an improvement
of several orders of magnitude in terms of prediction speed and a reduction in
error over previous estimators in various real-world data-sets.
|
1410.7540 | Swaleha Saeed | Swaleha Saeed, M Sarosh Umar, M Athar Ali and Musheer Ahmad | Fisher-Yates Chaotic Shuffling Based Image Encryption | null | International Journal of Information Processing, 8(3), 31-41, 2014
ISSN : 0973-8215 IK International Publishing House Pvt. Ltd., New Delhi,
India | null | null | cs.CR cs.MM | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In Present era, information security is of utmost concern and encryption is
one of the alternatives to ensure security. Chaos based cryptography has
brought a secure and efficient way to meet the challenges of secure multimedia
transmission over the networks. In this paper, we have proposed a secure
Grayscale image encryption methodology in wavelet domain. The proposed
algorithm performs shuffling followed by encryption using states of chaotic map
in a secure manner. Firstly, the image is transformed from spatial domain to
wavelet domain by the Haar wavelet. Subsequently, Fisher Yates chaotic
shuffling technique is employed to shuffle the image in wavelet domain to
confuse the relationship between plain image and cipher image. A key dependent
piece-wise linear chaotic map is used to generate chaos for the chaotic
shuffling. Further, the resultant shuffled approximate coefficients are
chaotically modulated. To enhance the statistical characteristics from
cryptographic point of view, the shuffled image is self keyed diffused and
mixing operation is carried out using keystream extracted from one-dimensional
chaotic map and the plain-image. The proposed algorithm is tested over some
standard image dataset. The results of several experimental, statistical and
sensitivity analyses proved that the algorithm provides an efficient and secure
method to achieve trusted gray scale image encryption.
| [
{
"version": "v1",
"created": "Tue, 28 Oct 2014 07:48:03 GMT"
}
] | 2014-10-29T00:00:00 | [
[
"Saeed",
"Swaleha",
""
],
[
"Umar",
"M Sarosh",
""
],
[
"Ali",
"M Athar",
""
],
[
"Ahmad",
"Musheer",
""
]
] | TITLE: Fisher-Yates Chaotic Shuffling Based Image Encryption
ABSTRACT: In Present era, information security is of utmost concern and encryption is
one of the alternatives to ensure security. Chaos based cryptography has
brought a secure and efficient way to meet the challenges of secure multimedia
transmission over the networks. In this paper, we have proposed a secure
Grayscale image encryption methodology in wavelet domain. The proposed
algorithm performs shuffling followed by encryption using states of chaotic map
in a secure manner. Firstly, the image is transformed from spatial domain to
wavelet domain by the Haar wavelet. Subsequently, Fisher Yates chaotic
shuffling technique is employed to shuffle the image in wavelet domain to
confuse the relationship between plain image and cipher image. A key dependent
piece-wise linear chaotic map is used to generate chaos for the chaotic
shuffling. Further, the resultant shuffled approximate coefficients are
chaotically modulated. To enhance the statistical characteristics from
cryptographic point of view, the shuffled image is self keyed diffused and
mixing operation is carried out using keystream extracted from one-dimensional
chaotic map and the plain-image. The proposed algorithm is tested over some
standard image dataset. The results of several experimental, statistical and
sensitivity analyses proved that the algorithm provides an efficient and secure
method to achieve trusted gray scale image encryption.
|
1410.7744 | Vincent Primault | Vincent Primault, Sonia Ben Mokhtar, Cedric Lauradoux, Lionel Brunie | Differentially Private Location Privacy in Practice | In Proceedings of the Third Workshop on Mobile Security Technologies
(MoST) 2014 (http://arxiv.org/abs/1410.6674) | null | null | MoST/2014/02 | cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | With the wide adoption of handheld devices (e.g. smartphones, tablets) a
large number of location-based services (also called LBSs) have flourished
providing mobile users with real-time and contextual information on the move.
Accounting for the amount of location information they are given by users,
these services are able to track users wherever they go and to learn sensitive
information about them (e.g. their points of interest including home, work,
religious or political places regularly visited). A number of solutions have
been proposed in the past few years to protect users location information while
still allowing them to enjoy geo-located services. Among the most robust
solutions are those that apply the popular notion of differential privacy to
location privacy (e.g. Geo-Indistinguishability), promising strong theoretical
privacy guarantees with a bounded accuracy loss. While these theoretical
guarantees are attracting, it might be difficult for end users or practitioners
to assess their effectiveness in the wild. In this paper, we carry on a
practical study using real mobility traces coming from two different datasets,
to assess the ability of Geo-Indistinguishability to protect users' points of
interest (POIs). We show that a curious LBS collecting obfuscated location
information sent by mobile users is still able to infer most of the users POIs
with a reasonable both geographic and semantic precision. This precision
depends on the degree of obfuscation applied by Geo-Indistinguishability.
Nevertheless, the latter also has an impact on the overhead incurred on mobile
devices resulting in a privacy versus overhead trade-off. Finally, we show in
our study that POIs constitute a quasi-identifier for mobile users and that
obfuscating them using Geo-Indistinguishability is not sufficient as an
attacker is able to re-identify at least 63% of them despite a high degree of
obfuscation.
| [
{
"version": "v1",
"created": "Tue, 28 Oct 2014 19:18:31 GMT"
}
] | 2014-10-29T00:00:00 | [
[
"Primault",
"Vincent",
""
],
[
"Mokhtar",
"Sonia Ben",
""
],
[
"Lauradoux",
"Cedric",
""
],
[
"Brunie",
"Lionel",
""
]
] | TITLE: Differentially Private Location Privacy in Practice
ABSTRACT: With the wide adoption of handheld devices (e.g. smartphones, tablets) a
large number of location-based services (also called LBSs) have flourished
providing mobile users with real-time and contextual information on the move.
Accounting for the amount of location information they are given by users,
these services are able to track users wherever they go and to learn sensitive
information about them (e.g. their points of interest including home, work,
religious or political places regularly visited). A number of solutions have
been proposed in the past few years to protect users location information while
still allowing them to enjoy geo-located services. Among the most robust
solutions are those that apply the popular notion of differential privacy to
location privacy (e.g. Geo-Indistinguishability), promising strong theoretical
privacy guarantees with a bounded accuracy loss. While these theoretical
guarantees are attracting, it might be difficult for end users or practitioners
to assess their effectiveness in the wild. In this paper, we carry on a
practical study using real mobility traces coming from two different datasets,
to assess the ability of Geo-Indistinguishability to protect users' points of
interest (POIs). We show that a curious LBS collecting obfuscated location
information sent by mobile users is still able to infer most of the users POIs
with a reasonable both geographic and semantic precision. This precision
depends on the degree of obfuscation applied by Geo-Indistinguishability.
Nevertheless, the latter also has an impact on the overhead incurred on mobile
devices resulting in a privacy versus overhead trade-off. Finally, we show in
our study that POIs constitute a quasi-identifier for mobile users and that
obfuscating them using Geo-Indistinguishability is not sufficient as an
attacker is able to re-identify at least 63% of them despite a high degree of
obfuscation.
|
1410.7758 | Tobias Blanke | Tobias Blanke, Mark Hedges | Towards a Virtual Data Centre for Classics | null | null | null | null | cs.DL | http://creativecommons.org/licenses/by-nc-sa/3.0/ | The paper presents some of our work on integrating datasets in Classics. We
present the results of various projects we had in this domain. The conclusions
from LaQuAT concerned limitations to the approach rather than solutions. The
relational model followed by OGSA-DAI was more effective for resources that
consist primarily of structured data (which we call data-centric) rather than
for largely unstructured text (which we call text-centric), which makes up a
significant component of the datasets we were using. This approach was,
moreover, insufficiently flexible to deal with the semantic issues. The gMan
project, on the other hand, addressed these problems by virtualizing data
resources using full-text indexes, which can then be used to provide different
views onto the collections and services that more closely match the sort of
information organization and retrieval activities found in the humanities, in
an environment that is more interactive, researcher-focused, and
researcher-driven.
| [
{
"version": "v1",
"created": "Tue, 28 Oct 2014 19:25:52 GMT"
}
] | 2014-10-29T00:00:00 | [
[
"Blanke",
"Tobias",
""
],
[
"Hedges",
"Mark",
""
]
] | TITLE: Towards a Virtual Data Centre for Classics
ABSTRACT: The paper presents some of our work on integrating datasets in Classics. We
present the results of various projects we had in this domain. The conclusions
from LaQuAT concerned limitations to the approach rather than solutions. The
relational model followed by OGSA-DAI was more effective for resources that
consist primarily of structured data (which we call data-centric) rather than
for largely unstructured text (which we call text-centric), which makes up a
significant component of the datasets we were using. This approach was,
moreover, insufficiently flexible to deal with the semantic issues. The gMan
project, on the other hand, addressed these problems by virtualizing data
resources using full-text indexes, which can then be used to provide different
views onto the collections and services that more closely match the sort of
information organization and retrieval activities found in the humanities, in
an environment that is more interactive, researcher-focused, and
researcher-driven.
|
1109.5460 | Ke Xu | Xiao Liang, Xudong Zheng, Weifeng Lv, Tongyu Zhu, Ke Xu | The scaling of human mobility by taxis is exponential | 20 pages, 7 figures | Physica A 391 (2012) 2135-2144 | 10.1016/j.physa.2011.11.035 | null | physics.soc-ph cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | As a significant factor in urban planning, traffic forecasting and prediction
of epidemics, modeling patterns of human mobility draws intensive attention
from researchers for decades. Power-law distribution and its variations are
observed from quite a few real-world human mobility datasets such as the
movements of banking notes, trackings of cell phone users' locations and
trajectories of vehicles. In this paper, we build models for 20 million
trajectories with fine granularity collected from more than 10 thousand taxis
in Beijing. In contrast to most models observed in human mobility data, the
taxis' traveling displacements in urban areas tend to follow an exponential
distribution instead of a power-law. Similarly, the elapsed time can also be
well approximated by an exponential distribution. Worth mentioning, analysis of
the interevent time indicates the bursty nature of human mobility, similar to
many other human activities.
| [
{
"version": "v1",
"created": "Mon, 26 Sep 2011 07:20:32 GMT"
}
] | 2014-10-28T00:00:00 | [
[
"Liang",
"Xiao",
""
],
[
"Zheng",
"Xudong",
""
],
[
"Lv",
"Weifeng",
""
],
[
"Zhu",
"Tongyu",
""
],
[
"Xu",
"Ke",
""
]
] | TITLE: The scaling of human mobility by taxis is exponential
ABSTRACT: As a significant factor in urban planning, traffic forecasting and prediction
of epidemics, modeling patterns of human mobility draws intensive attention
from researchers for decades. Power-law distribution and its variations are
observed from quite a few real-world human mobility datasets such as the
movements of banking notes, trackings of cell phone users' locations and
trajectories of vehicles. In this paper, we build models for 20 million
trajectories with fine granularity collected from more than 10 thousand taxis
in Beijing. In contrast to most models observed in human mobility data, the
taxis' traveling displacements in urban areas tend to follow an exponential
distribution instead of a power-law. Similarly, the elapsed time can also be
well approximated by an exponential distribution. Worth mentioning, analysis of
the interevent time indicates the bursty nature of human mobility, similar to
many other human activities.
|
1311.2887 | Faraz Zaidi | Aneeq Hashmi, Faraz Zaidi, Arnaud Sallaberry, Tariq Mehmood | Are all Social Networks Structurally Similar? A Comparative Study using
Network Statistics and Metrics | ASONAM 2012, Istanbul : Turkey (2012) | null | null | null | cs.SI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The modern age has seen an exponential growth of social network data
available on the web. Analysis of these networks reveal important structural
information about these networks in particular and about our societies in
general. More often than not, analysis of these networks is concerned in
identifying similarities among social networks and how they are different from
other networks such as protein interaction networks, computer networks and food
web. In this paper, our objective is to perform a critical analysis of
different social networks using structural metrics in an effort to highlight
their similarities and differences. We use five different social network
datasets which are contextually and semantically different from each other. We
then analyze these networks using a number of different network statistics and
metrics. Our results show that although these social networks have been
constructed from different contexts, they are structurally similar. We also
review the snowball sampling method and show its vulnerability against
different network metrics.
| [
{
"version": "v1",
"created": "Wed, 30 Oct 2013 09:55:12 GMT"
},
{
"version": "v2",
"created": "Mon, 27 Oct 2014 00:08:13 GMT"
}
] | 2014-10-28T00:00:00 | [
[
"Hashmi",
"Aneeq",
""
],
[
"Zaidi",
"Faraz",
""
],
[
"Sallaberry",
"Arnaud",
""
],
[
"Mehmood",
"Tariq",
""
]
] | TITLE: Are all Social Networks Structurally Similar? A Comparative Study using
Network Statistics and Metrics
ABSTRACT: The modern age has seen an exponential growth of social network data
available on the web. Analysis of these networks reveal important structural
information about these networks in particular and about our societies in
general. More often than not, analysis of these networks is concerned in
identifying similarities among social networks and how they are different from
other networks such as protein interaction networks, computer networks and food
web. In this paper, our objective is to perform a critical analysis of
different social networks using structural metrics in an effort to highlight
their similarities and differences. We use five different social network
datasets which are contextually and semantically different from each other. We
then analyze these networks using a number of different network statistics and
metrics. Our results show that although these social networks have been
constructed from different contexts, they are structurally similar. We also
review the snowball sampling method and show its vulnerability against
different network metrics.
|
1404.7206 | Qinsi Wang | Qinsi Wang, Paolo Zuliani, Soonho Kong, Sicun Gao, and Edmund M.
Clarke | SReach: A Bounded Model Checker for Stochastic Hybrid Systems | null | null | null | null | cs.FL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper we describe a new tool, SReach, which solves probabilistic
bounded reachability problems for two classes of stochastic hybrid systems. The
first one is (nonlinear) hybrid automata with parametric uncertainty. The
second one is probabilistic hybrid automata with additional randomness for both
transition probabilities and variable resets. Standard approaches to
reachability problems for linear hybrid systems require numerical solutions for
large optimization problems, and become infeasible for systems involving both
nonlinear dynamics over the reals and stochasticity. Our approach encodes
stochastic information by using random variables, and combines the randomized
sampling, a $\delta$-complete decision procedure, and statistical tests. SReach
utilizes the $\delta$-complete decision procedure to solve reachability
problems in a sound manner, i.e., it always decides correctly if, for a given
assignment to all random variables, the system actually reaches the unsafe
region. The statistical tests adapted guarantee arbitrary small error bounds
between probabilities estimated by SReach and real ones. Compared to standard
simulation-based methods, our approach supports non-deterministic branching,
increases the coverage of simulation, and avoids the zero-crossing problem. We
demonstrate our method's feasibility by applying SReach to three representative
biological models and to additional benchmarks for nonlinear hybrid systems
with multiple probabilistic system parameters.
| [
{
"version": "v1",
"created": "Tue, 29 Apr 2014 01:18:14 GMT"
},
{
"version": "v2",
"created": "Mon, 27 Oct 2014 15:36:45 GMT"
}
] | 2014-10-28T00:00:00 | [
[
"Wang",
"Qinsi",
""
],
[
"Zuliani",
"Paolo",
""
],
[
"Kong",
"Soonho",
""
],
[
"Gao",
"Sicun",
""
],
[
"Clarke",
"Edmund M.",
""
]
] | TITLE: SReach: A Bounded Model Checker for Stochastic Hybrid Systems
ABSTRACT: In this paper we describe a new tool, SReach, which solves probabilistic
bounded reachability problems for two classes of stochastic hybrid systems. The
first one is (nonlinear) hybrid automata with parametric uncertainty. The
second one is probabilistic hybrid automata with additional randomness for both
transition probabilities and variable resets. Standard approaches to
reachability problems for linear hybrid systems require numerical solutions for
large optimization problems, and become infeasible for systems involving both
nonlinear dynamics over the reals and stochasticity. Our approach encodes
stochastic information by using random variables, and combines the randomized
sampling, a $\delta$-complete decision procedure, and statistical tests. SReach
utilizes the $\delta$-complete decision procedure to solve reachability
problems in a sound manner, i.e., it always decides correctly if, for a given
assignment to all random variables, the system actually reaches the unsafe
region. The statistical tests adapted guarantee arbitrary small error bounds
between probabilities estimated by SReach and real ones. Compared to standard
simulation-based methods, our approach supports non-deterministic branching,
increases the coverage of simulation, and avoids the zero-crossing problem. We
demonstrate our method's feasibility by applying SReach to three representative
biological models and to additional benchmarks for nonlinear hybrid systems
with multiple probabilistic system parameters.
|
1410.6824 | Ramy Medhat | Ramy Medhat, Borzoo Bonakdarpour, Sebastian Fischmeister | Power Redistribution for Optimizing Performance in MPI Clusters | null | null | null | null | cs.DC | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Power efficiency has recently become a major concern in the high-performance
computing domain. HPC centers are provisioned by a power bound which impacts
execution time. Naturally, a tradeoff arises between power efficiency and
computational efficiency. This paper tackles the problem of performance
optimization for MPI applications, where a power bound is assumed. The paper
exposes a subset of HPC applications that leverage cluster parallelism using
MPI, where nodes encounter multiple synchronization points and exhibit
inter-node dependency. We abstract this structure into a dependency graph, and
leverage the asymmetry in execution time of parallel jobs on different nodes by
redistributing power gained from idling a blocked node to nodes that are
lagging in their jobs. We introduce a solution based on integer linear
programming (ILP) for optimal power distribution algorithm that minimizes total
execution time, while maintaining an upper power bound. We then present an
online heuristic that dynamically redistributes power at run time. The
heuristic shows significant reductions in total execution time of a set of
parallel benchmarks with speedup up to 2.25x.
| [
{
"version": "v1",
"created": "Fri, 24 Oct 2014 20:08:33 GMT"
}
] | 2014-10-28T00:00:00 | [
[
"Medhat",
"Ramy",
""
],
[
"Bonakdarpour",
"Borzoo",
""
],
[
"Fischmeister",
"Sebastian",
""
]
] | TITLE: Power Redistribution for Optimizing Performance in MPI Clusters
ABSTRACT: Power efficiency has recently become a major concern in the high-performance
computing domain. HPC centers are provisioned by a power bound which impacts
execution time. Naturally, a tradeoff arises between power efficiency and
computational efficiency. This paper tackles the problem of performance
optimization for MPI applications, where a power bound is assumed. The paper
exposes a subset of HPC applications that leverage cluster parallelism using
MPI, where nodes encounter multiple synchronization points and exhibit
inter-node dependency. We abstract this structure into a dependency graph, and
leverage the asymmetry in execution time of parallel jobs on different nodes by
redistributing power gained from idling a blocked node to nodes that are
lagging in their jobs. We introduce a solution based on integer linear
programming (ILP) for optimal power distribution algorithm that minimizes total
execution time, while maintaining an upper power bound. We then present an
online heuristic that dynamically redistributes power at run time. The
heuristic shows significant reductions in total execution time of a set of
parallel benchmarks with speedup up to 2.25x.
|
1410.6880 | Seunghak Lee | Seunghak Lee and Eric P. Xing | Screening Rules for Overlapping Group Lasso | null | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Recently, to solve large-scale lasso and group lasso problems, screening
rules have been developed, the goal of which is to reduce the problem size by
efficiently discarding zero coefficients using simple rules independently of
the others. However, screening for overlapping group lasso remains an open
challenge because the overlaps between groups make it infeasible to test each
group independently. In this paper, we develop screening rules for overlapping
group lasso. To address the challenge arising from groups with overlaps, we
take into account overlapping groups only if they are inclusive of the group
being tested, and then we derive screening rules, adopting the dual polytope
projection approach. This strategy allows us to screen each group independently
of each other. In our experiments, we demonstrate the efficiency of our
screening rules on various datasets.
| [
{
"version": "v1",
"created": "Sat, 25 Oct 2014 04:06:49 GMT"
}
] | 2014-10-28T00:00:00 | [
[
"Lee",
"Seunghak",
""
],
[
"Xing",
"Eric P.",
""
]
] | TITLE: Screening Rules for Overlapping Group Lasso
ABSTRACT: Recently, to solve large-scale lasso and group lasso problems, screening
rules have been developed, the goal of which is to reduce the problem size by
efficiently discarding zero coefficients using simple rules independently of
the others. However, screening for overlapping group lasso remains an open
challenge because the overlaps between groups make it infeasible to test each
group independently. In this paper, we develop screening rules for overlapping
group lasso. To address the challenge arising from groups with overlaps, we
take into account overlapping groups only if they are inclusive of the group
being tested, and then we derive screening rules, adopting the dual polytope
projection approach. This strategy allows us to screen each group independently
of each other. In our experiments, we demonstrate the efficiency of our
screening rules on various datasets.
|
1410.6990 | Dacheng Tao | Chang Xu, Tongliang Liu, Dacheng Tao, Chao Xu | Local Rademacher Complexity for Multi-label Learning | null | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We analyze the local Rademacher complexity of empirical risk minimization
(ERM)-based multi-label learning algorithms, and in doing so propose a new
algorithm for multi-label learning. Rather than using the trace norm to
regularize the multi-label predictor, we instead minimize the tail sum of the
singular values of the predictor in multi-label learning. Benefiting from the
use of the local Rademacher complexity, our algorithm, therefore, has a sharper
generalization error bound and a faster convergence rate. Compared to methods
that minimize over all singular values, concentrating on the tail singular
values results in better recovery of the low-rank structure of the multi-label
predictor, which plays an import role in exploiting label correlations. We
propose a new conditional singular value thresholding algorithm to solve the
resulting objective function. Empirical studies on real-world datasets validate
our theoretical results and demonstrate the effectiveness of the proposed
algorithm.
| [
{
"version": "v1",
"created": "Sun, 26 Oct 2014 05:52:33 GMT"
}
] | 2014-10-28T00:00:00 | [
[
"Xu",
"Chang",
""
],
[
"Liu",
"Tongliang",
""
],
[
"Tao",
"Dacheng",
""
],
[
"Xu",
"Chao",
""
]
] | TITLE: Local Rademacher Complexity for Multi-label Learning
ABSTRACT: We analyze the local Rademacher complexity of empirical risk minimization
(ERM)-based multi-label learning algorithms, and in doing so propose a new
algorithm for multi-label learning. Rather than using the trace norm to
regularize the multi-label predictor, we instead minimize the tail sum of the
singular values of the predictor in multi-label learning. Benefiting from the
use of the local Rademacher complexity, our algorithm, therefore, has a sharper
generalization error bound and a faster convergence rate. Compared to methods
that minimize over all singular values, concentrating on the tail singular
values results in better recovery of the low-rank structure of the multi-label
predictor, which plays an import role in exploiting label correlations. We
propose a new conditional singular value thresholding algorithm to solve the
resulting objective function. Empirical studies on real-world datasets validate
our theoretical results and demonstrate the effectiveness of the proposed
algorithm.
|
1410.6996 | Musa Maharramov | Musa Maharramov and Biondo Biondi | Improved depth imaging by constrained full-waveform inversion | 5 pages, 2 figures | null | null | SEP 155 | physics.geo-ph cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We propose a formulation of full-wavefield inversion (FWI) as a constrained
optimization problem, and describe a computationally efficient technique for
solving constrained full-wavefield inversion (CFWI). The technique is based on
using a total-variation regularization method, with the regularization weighted
in favor of constraining deeper subsurface model sections. The method helps to
promote "edge-preserving" blocky model inversion where fitting the seismic data
alone fails to adequately constrain the model. The method is demonstrated on
synthetic datasets with added noise, and is shown to enhance the sharpness of
the inverted model and correctly reposition mispositioned reflectors by better
constraining the velocity model at depth.
| [
{
"version": "v1",
"created": "Sun, 26 Oct 2014 08:02:01 GMT"
}
] | 2014-10-28T00:00:00 | [
[
"Maharramov",
"Musa",
""
],
[
"Biondi",
"Biondo",
""
]
] | TITLE: Improved depth imaging by constrained full-waveform inversion
ABSTRACT: We propose a formulation of full-wavefield inversion (FWI) as a constrained
optimization problem, and describe a computationally efficient technique for
solving constrained full-wavefield inversion (CFWI). The technique is based on
using a total-variation regularization method, with the regularization weighted
in favor of constraining deeper subsurface model sections. The method helps to
promote "edge-preserving" blocky model inversion where fitting the seismic data
alone fails to adequately constrain the model. The method is demonstrated on
synthetic datasets with added noise, and is shown to enhance the sharpness of
the inverted model and correctly reposition mispositioned reflectors by better
constraining the velocity model at depth.
|
1410.7100 | Harris Georgiou | Harris V. Georgiou | Estimating the intrinsic dimension in fMRI space via dataset fractal
analysis - Counting the `cpu cores' of the human brain | 27 pages, 10 figures, 2 tables, 47 references | null | null | HG/AI.1014.27v1 (draft/preprint) | cs.AI cs.CV q-bio.NC stat.ML | http://creativecommons.org/licenses/by-nc-sa/3.0/ | Functional Magnetic Resonance Imaging (fMRI) is a powerful non-invasive tool
for localizing and analyzing brain activity. This study focuses on one very
important aspect of the functional properties of human brain, specifically the
estimation of the level of parallelism when performing complex cognitive tasks.
Using fMRI as the main modality, the human brain activity is investigated
through a purely data-driven signal processing and dimensionality analysis
approach. Specifically, the fMRI signal is treated as a multi-dimensional data
space and its intrinsic `complexity' is studied via dataset fractal analysis
and blind-source separation (BSS) methods. One simulated and two real fMRI
datasets are used in combination with Independent Component Analysis (ICA) and
fractal analysis for estimating the intrinsic (true) dimensionality, in order
to provide data-driven experimental evidence on the number of independent brain
processes that run in parallel when visual or visuo-motor tasks are performed.
Although this number is can not be defined as a strict threshold but rather as
a continuous range, when a specific activation level is defined, a
corresponding number of parallel processes or the casual equivalent of `cpu
cores' can be detected in normal human brain activity.
| [
{
"version": "v1",
"created": "Mon, 27 Oct 2014 00:25:24 GMT"
}
] | 2014-10-28T00:00:00 | [
[
"Georgiou",
"Harris V.",
""
]
] | TITLE: Estimating the intrinsic dimension in fMRI space via dataset fractal
analysis - Counting the `cpu cores' of the human brain
ABSTRACT: Functional Magnetic Resonance Imaging (fMRI) is a powerful non-invasive tool
for localizing and analyzing brain activity. This study focuses on one very
important aspect of the functional properties of human brain, specifically the
estimation of the level of parallelism when performing complex cognitive tasks.
Using fMRI as the main modality, the human brain activity is investigated
through a purely data-driven signal processing and dimensionality analysis
approach. Specifically, the fMRI signal is treated as a multi-dimensional data
space and its intrinsic `complexity' is studied via dataset fractal analysis
and blind-source separation (BSS) methods. One simulated and two real fMRI
datasets are used in combination with Independent Component Analysis (ICA) and
fractal analysis for estimating the intrinsic (true) dimensionality, in order
to provide data-driven experimental evidence on the number of independent brain
processes that run in parallel when visual or visuo-motor tasks are performed.
Although this number is can not be defined as a strict threshold but rather as
a continuous range, when a specific activation level is defined, a
corresponding number of parallel processes or the casual equivalent of `cpu
cores' can be detected in normal human brain activity.
|
1410.7372 | Jayadeva | Jayadeva, Sanjit S. Batra, and Siddharth Sabharwal | Feature Selection through Minimization of the VC dimension | arXiv admin note: text overlap with arXiv:1410.4573 | null | null | null | cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Feature selection involes identifying the most relevant subset of input
features, with a view to improving generalization of predictive models by
reducing overfitting. Directly searching for the most relevant combination of
attributes is NP-hard. Variable selection is of critical importance in many
applications, such as micro-array data analysis, where selecting a small number
of discriminative features is crucial to developing useful models of disease
mechanisms, as well as for prioritizing targets for drug discovery. The
recently proposed Minimal Complexity Machine (MCM) provides a way to learn a
hyperplane classifier by minimizing an exact (\boldmath{$\Theta$}) bound on its
VC dimension. It is well known that a lower VC dimension contributes to good
generalization. For a linear hyperplane classifier in the input space, the VC
dimension is upper bounded by the number of features; hence, a linear
classifier with a small VC dimension is parsimonious in the set of features it
employs. In this paper, we use the linear MCM to learn a classifier in which a
large number of weights are zero; features with non-zero weights are the ones
that are chosen. Selected features are used to learn a kernel SVM classifier.
On a number of benchmark datasets, the features chosen by the linear MCM yield
comparable or better test set accuracy than when methods such as ReliefF and
FCBF are used for the task. The linear MCM typically chooses one-tenth the
number of attributes chosen by the other methods; on some very high dimensional
datasets, the MCM chooses about $0.6\%$ of the features; in comparison, ReliefF
and FCBF choose 70 to 140 times more features, thus demonstrating that
minimizing the VC dimension may provide a new, and very effective route for
feature selection and for learning sparse representations.
| [
{
"version": "v1",
"created": "Mon, 27 Oct 2014 19:46:55 GMT"
}
] | 2014-10-28T00:00:00 | [
[
"Jayadeva",
"",
""
],
[
"Batra",
"Sanjit S.",
""
],
[
"Sabharwal",
"Siddharth",
""
]
] | TITLE: Feature Selection through Minimization of the VC dimension
ABSTRACT: Feature selection involes identifying the most relevant subset of input
features, with a view to improving generalization of predictive models by
reducing overfitting. Directly searching for the most relevant combination of
attributes is NP-hard. Variable selection is of critical importance in many
applications, such as micro-array data analysis, where selecting a small number
of discriminative features is crucial to developing useful models of disease
mechanisms, as well as for prioritizing targets for drug discovery. The
recently proposed Minimal Complexity Machine (MCM) provides a way to learn a
hyperplane classifier by minimizing an exact (\boldmath{$\Theta$}) bound on its
VC dimension. It is well known that a lower VC dimension contributes to good
generalization. For a linear hyperplane classifier in the input space, the VC
dimension is upper bounded by the number of features; hence, a linear
classifier with a small VC dimension is parsimonious in the set of features it
employs. In this paper, we use the linear MCM to learn a classifier in which a
large number of weights are zero; features with non-zero weights are the ones
that are chosen. Selected features are used to learn a kernel SVM classifier.
On a number of benchmark datasets, the features chosen by the linear MCM yield
comparable or better test set accuracy than when methods such as ReliefF and
FCBF are used for the task. The linear MCM typically chooses one-tenth the
number of attributes chosen by the other methods; on some very high dimensional
datasets, the MCM chooses about $0.6\%$ of the features; in comparison, ReliefF
and FCBF choose 70 to 140 times more features, thus demonstrating that
minimizing the VC dimension may provide a new, and very effective route for
feature selection and for learning sparse representations.
|
1410.6532 | Ziming Zhang | Ziming Zhang, Yuting Chen, Venkatesh Saligrama | A Novel Visual Word Co-occurrence Model for Person Re-identification | Accepted at ECCV Workshop on Visual Surveillance and
Re-Identification, 2014 | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Person re-identification aims to maintain the identity of an individual in
diverse locations through different non-overlapping camera views. The problem
is fundamentally challenging due to appearance variations resulting from
differing poses, illumination and configurations of camera views. To deal with
these difficulties, we propose a novel visual word co-occurrence model. We
first map each pixel of an image to a visual word using a codebook, which is
learned in an unsupervised manner. The appearance transformation between camera
views is encoded by a co-occurrence matrix of visual word joint distributions
in probe and gallery images. Our appearance model naturally accounts for
spatial similarities and variations caused by pose, illumination &
configuration change across camera views. Linear SVMs are then trained as
classifiers using these co-occurrence descriptors. On the VIPeR and CUHK Campus
benchmark datasets, our method achieves 83.86% and 85.49% at rank-15 on the
Cumulative Match Characteristic (CMC) curves, and beats the state-of-the-art
results by 10.44% and 22.27%.
| [
{
"version": "v1",
"created": "Fri, 24 Oct 2014 01:04:37 GMT"
}
] | 2014-10-27T00:00:00 | [
[
"Zhang",
"Ziming",
""
],
[
"Chen",
"Yuting",
""
],
[
"Saligrama",
"Venkatesh",
""
]
] | TITLE: A Novel Visual Word Co-occurrence Model for Person Re-identification
ABSTRACT: Person re-identification aims to maintain the identity of an individual in
diverse locations through different non-overlapping camera views. The problem
is fundamentally challenging due to appearance variations resulting from
differing poses, illumination and configurations of camera views. To deal with
these difficulties, we propose a novel visual word co-occurrence model. We
first map each pixel of an image to a visual word using a codebook, which is
learned in an unsupervised manner. The appearance transformation between camera
views is encoded by a co-occurrence matrix of visual word joint distributions
in probe and gallery images. Our appearance model naturally accounts for
spatial similarities and variations caused by pose, illumination &
configuration change across camera views. Linear SVMs are then trained as
classifiers using these co-occurrence descriptors. On the VIPeR and CUHK Campus
benchmark datasets, our method achieves 83.86% and 85.49% at rank-15 on the
Cumulative Match Characteristic (CMC) curves, and beats the state-of-the-art
results by 10.44% and 22.27%.
|
1410.6609 | Dominik Bartuschat | Dominik Bartuschat, Ulrich R\"ude | Parallel Multiphysics Simulations of Charged Particles in Microfluidic
Flows | Submitted to Journal of Computational Science (Elsevier) | null | null | null | cs.CE physics.comp-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The article describes parallel multiphysics simulations of charged particles
in microfluidic flows with the waLBerla framework. To this end, three physical
effects are coupled: rigid body dynamics, fluid flow modelled by a lattice
Boltzmann algorithm, and electric potentials represented by a finite volume
discretisation. For solving the finite volume discretisation for the
electrostatic forces, a cell-centered multigrid algorithm is developed that
conforms to the lattice Boltzmann meshes and the parallel communication
structure of waLBerla. The new functionality is validated with suitable
benchmark scenarios. Additionally, the parallel scaling and the numerical
efficiency of the algorithms are analysed on an advanced supercomputer.
| [
{
"version": "v1",
"created": "Fri, 24 Oct 2014 08:10:13 GMT"
}
] | 2014-10-27T00:00:00 | [
[
"Bartuschat",
"Dominik",
""
],
[
"Rüde",
"Ulrich",
""
]
] | TITLE: Parallel Multiphysics Simulations of Charged Particles in Microfluidic
Flows
ABSTRACT: The article describes parallel multiphysics simulations of charged particles
in microfluidic flows with the waLBerla framework. To this end, three physical
effects are coupled: rigid body dynamics, fluid flow modelled by a lattice
Boltzmann algorithm, and electric potentials represented by a finite volume
discretisation. For solving the finite volume discretisation for the
electrostatic forces, a cell-centered multigrid algorithm is developed that
conforms to the lattice Boltzmann meshes and the parallel communication
structure of waLBerla. The new functionality is validated with suitable
benchmark scenarios. Additionally, the parallel scaling and the numerical
efficiency of the algorithms are analysed on an advanced supercomputer.
|
1410.6629 | Gianluca Stringhini | Gianluca Stringhini, Olivier Thonnard | That Ain't You: Detecting Spearphishing Emails Before They Are Sent | null | null | null | null | cs.CR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | One of the ways in which attackers try to steal sensitive information from
corporations is by sending spearphishing emails. This type of emails typically
appear to be sent by one of the victim's coworkers, but have instead been
crafted by an attacker. A particularly insidious type of spearphishing emails
are the ones that do not only claim to come from a trusted party, but were
actually sent from that party's legitimate email account that was compromised
in the first place. In this paper, we propose a radical change of focus in the
techniques used for detecting such malicious emails: instead of looking for
particular features that are indicative of attack emails, we look for possible
indicators of impersonation of the legitimate owners. We present
IdentityMailer, a system that validates the authorship of emails by learning
the typical email-sending behavior of users over time, and comparing any
subsequent email sent from their accounts against this model. Our experiments
on real world e-mail datasets demonstrate that our system can effectively block
advanced email attacks sent from genuine email accounts, which traditional
protection systems are unable to detect. Moreover, we show that it is resilient
to an attacker willing to evade the system. To the best of our knowledge,
IdentityMailer is the first system able to identify spearphishing emails that
are sent from within an organization, by a skilled attacker having access to a
compromised email account.
| [
{
"version": "v1",
"created": "Fri, 24 Oct 2014 09:45:03 GMT"
}
] | 2014-10-27T00:00:00 | [
[
"Stringhini",
"Gianluca",
""
],
[
"Thonnard",
"Olivier",
""
]
] | TITLE: That Ain't You: Detecting Spearphishing Emails Before They Are Sent
ABSTRACT: One of the ways in which attackers try to steal sensitive information from
corporations is by sending spearphishing emails. This type of emails typically
appear to be sent by one of the victim's coworkers, but have instead been
crafted by an attacker. A particularly insidious type of spearphishing emails
are the ones that do not only claim to come from a trusted party, but were
actually sent from that party's legitimate email account that was compromised
in the first place. In this paper, we propose a radical change of focus in the
techniques used for detecting such malicious emails: instead of looking for
particular features that are indicative of attack emails, we look for possible
indicators of impersonation of the legitimate owners. We present
IdentityMailer, a system that validates the authorship of emails by learning
the typical email-sending behavior of users over time, and comparing any
subsequent email sent from their accounts against this model. Our experiments
on real world e-mail datasets demonstrate that our system can effectively block
advanced email attacks sent from genuine email accounts, which traditional
protection systems are unable to detect. Moreover, we show that it is resilient
to an attacker willing to evade the system. To the best of our knowledge,
IdentityMailer is the first system able to identify spearphishing emails that
are sent from within an organization, by a skilled attacker having access to a
compromised email account.
|
1410.6725 | Mark Taylor | Mark Taylor | Visualising Large Datasets in TOPCAT v4 | 4 pages, 2 figures, conference paper submitted to arXiv a year after
acceptance | Astronomical Data Anaylsis Softward and Systems XXIII. Proceedings
of a meeting held 29 September - 3 October 2013 at Waikoloa Beach Marriott,
Hawaii, USA. Edited by N. Manset and P. Forshay ASP conference series, vol.
485, 2014, p.257 | null | null | astro-ph.IM cs.GR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | TOPCAT is a widely used desktop application for manipulation of astronomical
catalogues and other tables, which has long provided fast interactive
visualisation features including 1, 2 and 3-d plots, multiple datasets, linked
views, color coding, transparency and more. In Version 4 a new plotting library
has been written from scratch to deliver new and enhanced visualisation
capabilities. This paper describes some of the considerations in the design and
implementation, particularly in regard to providing comprehensible interactive
visualisation for multi-million point datasets.
| [
{
"version": "v1",
"created": "Fri, 24 Oct 2014 16:13:11 GMT"
}
] | 2014-10-27T00:00:00 | [
[
"Taylor",
"Mark",
""
]
] | TITLE: Visualising Large Datasets in TOPCAT v4
ABSTRACT: TOPCAT is a widely used desktop application for manipulation of astronomical
catalogues and other tables, which has long provided fast interactive
visualisation features including 1, 2 and 3-d plots, multiple datasets, linked
views, color coding, transparency and more. In Version 4 a new plotting library
has been written from scratch to deliver new and enhanced visualisation
capabilities. This paper describes some of the considerations in the design and
implementation, particularly in regard to providing comprehensible interactive
visualisation for multi-million point datasets.
|
1410.6776 | Purushottam Kar | Purushottam Kar, Harikrishna Narasimhan, Prateek Jain | Online and Stochastic Gradient Methods for Non-decomposable Loss
Functions | 25 pages, 3 figures, To appear in the proceedings of the 28th Annual
Conference on Neural Information Processing Systems, NIPS 2014 | null | null | null | cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Modern applications in sensitive domains such as biometrics and medicine
frequently require the use of non-decomposable loss functions such as
precision@k, F-measure etc. Compared to point loss functions such as
hinge-loss, these offer much more fine grained control over prediction, but at
the same time present novel challenges in terms of algorithm design and
analysis. In this work we initiate a study of online learning techniques for
such non-decomposable loss functions with an aim to enable incremental learning
as well as design scalable solvers for batch problems. To this end, we propose
an online learning framework for such loss functions. Our model enjoys several
nice properties, chief amongst them being the existence of efficient online
learning algorithms with sublinear regret and online to batch conversion
bounds. Our model is a provable extension of existing online learning models
for point loss functions. We instantiate two popular losses, prec@k and pAUC,
in our model and prove sublinear regret bounds for both of them. Our proofs
require a novel structural lemma over ranked lists which may be of independent
interest. We then develop scalable stochastic gradient descent solvers for
non-decomposable loss functions. We show that for a large family of loss
functions satisfying a certain uniform convergence property (that includes
prec@k, pAUC, and F-measure), our methods provably converge to the empirical
risk minimizer. Such uniform convergence results were not known for these
losses and we establish these using novel proof techniques. We then use
extensive experimentation on real life and benchmark datasets to establish that
our method can be orders of magnitude faster than a recently proposed cutting
plane method.
| [
{
"version": "v1",
"created": "Fri, 24 Oct 2014 18:45:23 GMT"
}
] | 2014-10-27T00:00:00 | [
[
"Kar",
"Purushottam",
""
],
[
"Narasimhan",
"Harikrishna",
""
],
[
"Jain",
"Prateek",
""
]
] | TITLE: Online and Stochastic Gradient Methods for Non-decomposable Loss
Functions
ABSTRACT: Modern applications in sensitive domains such as biometrics and medicine
frequently require the use of non-decomposable loss functions such as
precision@k, F-measure etc. Compared to point loss functions such as
hinge-loss, these offer much more fine grained control over prediction, but at
the same time present novel challenges in terms of algorithm design and
analysis. In this work we initiate a study of online learning techniques for
such non-decomposable loss functions with an aim to enable incremental learning
as well as design scalable solvers for batch problems. To this end, we propose
an online learning framework for such loss functions. Our model enjoys several
nice properties, chief amongst them being the existence of efficient online
learning algorithms with sublinear regret and online to batch conversion
bounds. Our model is a provable extension of existing online learning models
for point loss functions. We instantiate two popular losses, prec@k and pAUC,
in our model and prove sublinear regret bounds for both of them. Our proofs
require a novel structural lemma over ranked lists which may be of independent
interest. We then develop scalable stochastic gradient descent solvers for
non-decomposable loss functions. We show that for a large family of loss
functions satisfying a certain uniform convergence property (that includes
prec@k, pAUC, and F-measure), our methods provably converge to the empirical
risk minimizer. Such uniform convergence results were not known for these
losses and we establish these using novel proof techniques. We then use
extensive experimentation on real life and benchmark datasets to establish that
our method can be orders of magnitude faster than a recently proposed cutting
plane method.
|
1302.2837 | Sebastian Nanz | Sebastian Nanz, Scott West, Kaue Soares da Silveira, Bertrand Meyer | Benchmarking Usability and Performance of Multicore Languages | null | Proceedings of the 7th ACM-IEEE International Symposium Empirical
Software Engineering and Measurement (ESEM'13), pages 183-192. IEEE, 2013 | 10.1109/ESEM.2013.10 | null | cs.DC cs.PL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Developers face a wide choice of programming languages and libraries
supporting multicore computing. Ever more diverse paradigms for expressing
parallelism and synchronization become available while their influence on
usability and performance remains largely unclear. This paper describes an
experiment comparing four markedly different approaches to parallel
programming: Chapel, Cilk, Go, and Threading Building Blocks (TBB). Each
language is used to implement sequential and parallel versions of six benchmark
programs. The implementations are then reviewed by notable experts in the
language, thereby obtaining reference versions for each language and benchmark.
The resulting pool of 96 implementations is used to compare the languages with
respect to source code size, coding time, execution time, and speedup. The
experiment uncovers strengths and weaknesses in all approaches, facilitating an
informed selection of a language under a particular set of requirements. The
expert review step furthermore highlights the importance of expert knowledge
when using modern parallel programming approaches.
| [
{
"version": "v1",
"created": "Tue, 12 Feb 2013 16:08:12 GMT"
},
{
"version": "v2",
"created": "Thu, 23 Oct 2014 10:42:56 GMT"
}
] | 2014-10-24T00:00:00 | [
[
"Nanz",
"Sebastian",
""
],
[
"West",
"Scott",
""
],
[
"da Silveira",
"Kaue Soares",
""
],
[
"Meyer",
"Bertrand",
""
]
] | TITLE: Benchmarking Usability and Performance of Multicore Languages
ABSTRACT: Developers face a wide choice of programming languages and libraries
supporting multicore computing. Ever more diverse paradigms for expressing
parallelism and synchronization become available while their influence on
usability and performance remains largely unclear. This paper describes an
experiment comparing four markedly different approaches to parallel
programming: Chapel, Cilk, Go, and Threading Building Blocks (TBB). Each
language is used to implement sequential and parallel versions of six benchmark
programs. The implementations are then reviewed by notable experts in the
language, thereby obtaining reference versions for each language and benchmark.
The resulting pool of 96 implementations is used to compare the languages with
respect to source code size, coding time, execution time, and speedup. The
experiment uncovers strengths and weaknesses in all approaches, facilitating an
informed selection of a language under a particular set of requirements. The
expert review step furthermore highlights the importance of expert knowledge
when using modern parallel programming approaches.
|
1308.0938 | Mischael Schill | Mischael Schill, Sebastian Nanz, and Bertrand Meyer | Handling Parallelism in a Concurrency Model | MUSEPAT 2013 | Proceedings of the 2013 International Conference on Multicore
Software Engineering, Performance, and Tools (MUSEPAT'13), volume 8063 of
Lecture Notes in Computer Science, pages 37-48. Springer, 2013 | 10.1007/978-3-642-39955-8_4 | null | cs.SE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Programming models for concurrency are optimized for dealing with
nondeterminism, for example to handle asynchronously arriving events. To shield
the developer from data race errors effectively, such models may prevent shared
access to data altogether. However, this restriction also makes them unsuitable
for applications that require data parallelism. We present a library-based
approach for permitting parallel access to arrays while preserving the safety
guarantees of the original model. When applied to SCOOP, an object-oriented
concurrency model, the approach exhibits a negligible performance overhead
compared to ordinary threaded implementations of two parallel benchmark
programs.
| [
{
"version": "v1",
"created": "Mon, 5 Aug 2013 11:13:59 GMT"
}
] | 2014-10-24T00:00:00 | [
[
"Schill",
"Mischael",
""
],
[
"Nanz",
"Sebastian",
""
],
[
"Meyer",
"Bertrand",
""
]
] | TITLE: Handling Parallelism in a Concurrency Model
ABSTRACT: Programming models for concurrency are optimized for dealing with
nondeterminism, for example to handle asynchronously arriving events. To shield
the developer from data race errors effectively, such models may prevent shared
access to data altogether. However, this restriction also makes them unsuitable
for applications that require data parallelism. We present a library-based
approach for permitting parallel access to arrays while preserving the safety
guarantees of the original model. When applied to SCOOP, an object-oriented
concurrency model, the approach exhibits a negligible performance overhead
compared to ordinary threaded implementations of two parallel benchmark
programs.
|
1311.0269 | Peter Elmer | David Abdurachmanov, Peter Elmer, Giulio Eulisse, Shahzad Muzaffar | Initial explorations of ARM processors for scientific computing | Submitted to proceedings of the 15th International Workshop on
Advanced Computing and Analysis Techniques in Physics Research (ACAT2013),
Beijing. arXiv admin note: text overlap with arXiv:1311.1001 | null | 10.1088/1742-6596/523/1/012009 | null | physics.comp-ph cs.DC cs.NA hep-ex | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Power efficiency is becoming an ever more important metric for both high
performance and high throughput computing. Over the course of next decade it is
expected that flops/watt will be a major driver for the evolution of computer
architecture. Servers with large numbers of ARM processors, already ubiquitous
in mobile computing, are a promising alternative to traditional x86-64
computing. We present the results of our initial investigations into the use of
ARM processors for scientific computing applications. In particular we report
the results from our work with a current generation ARMv7 development board to
explore ARM-specific issues regarding the software development environment,
operating system, performance benchmarks and issues for porting High Energy
Physics software.
| [
{
"version": "v1",
"created": "Fri, 1 Nov 2013 19:33:35 GMT"
},
{
"version": "v2",
"created": "Wed, 22 Jan 2014 09:23:39 GMT"
}
] | 2014-10-24T00:00:00 | [
[
"Abdurachmanov",
"David",
""
],
[
"Elmer",
"Peter",
""
],
[
"Eulisse",
"Giulio",
""
],
[
"Muzaffar",
"Shahzad",
""
]
] | TITLE: Initial explorations of ARM processors for scientific computing
ABSTRACT: Power efficiency is becoming an ever more important metric for both high
performance and high throughput computing. Over the course of next decade it is
expected that flops/watt will be a major driver for the evolution of computer
architecture. Servers with large numbers of ARM processors, already ubiquitous
in mobile computing, are a promising alternative to traditional x86-64
computing. We present the results of our initial investigations into the use of
ARM processors for scientific computing applications. In particular we report
the results from our work with a current generation ARMv7 development board to
explore ARM-specific issues regarding the software development environment,
operating system, performance benchmarks and issues for porting High Energy
Physics software.
|
1311.1001 | Peter Elmer | David Abdurachmanov, Kapil Arya, Josh Bendavid, Tommaso Boccali, Gene
Cooperman, Andrea Dotti, Peter Elmer, Giulio Eulisse, Francesco Giacomini,
Christopher D. Jones, Matteo Manzali, Shahzad Muzaffar | Explorations of the viability of ARM and Xeon Phi for physics processing | Submitted to proceedings of the 20th International Conference on
Computing in High Energy and Nuclear Physics (CHEP13), Amsterdam | null | 10.1088/1742-6596/513/5/052008 | null | physics.comp-ph cs.DC hep-ex | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We report on our investigations into the viability of the ARM processor and
the Intel Xeon Phi co-processor for scientific computing. We describe our
experience porting software to these processors and running benchmarks using
real physics applications to explore the potential of these processors for
production physics processing.
| [
{
"version": "v1",
"created": "Tue, 5 Nov 2013 09:54:14 GMT"
},
{
"version": "v2",
"created": "Tue, 21 Jan 2014 18:38:34 GMT"
}
] | 2014-10-24T00:00:00 | [
[
"Abdurachmanov",
"David",
""
],
[
"Arya",
"Kapil",
""
],
[
"Bendavid",
"Josh",
""
],
[
"Boccali",
"Tommaso",
""
],
[
"Cooperman",
"Gene",
""
],
[
"Dotti",
"Andrea",
""
],
[
"Elmer",
"Peter",
""
],
[
"Eulisse",
"Giulio",
""
],
[
"Giacomini",
"Francesco",
""
],
[
"Jones",
"Christopher D.",
""
],
[
"Manzali",
"Matteo",
""
],
[
"Muzaffar",
"Shahzad",
""
]
] | TITLE: Explorations of the viability of ARM and Xeon Phi for physics processing
ABSTRACT: We report on our investigations into the viability of the ARM processor and
the Intel Xeon Phi co-processor for scientific computing. We describe our
experience porting software to these processors and running benchmarks using
real physics applications to explore the potential of these processors for
production physics processing.
|
1404.4774 | Jing Wang | Wang Jing, Zhao Zhong-Qiu, Hu Xuegang, Cheung Yiu-ming, Wang Meng, Wu
Xindong | Online Group Feature Selection | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Online feature selection with dynamic features has become an active research
area in recent years. However, in some real-world applications such as image
analysis and email spam filtering, features may arrive by groups. Existing
online feature selection methods evaluate features individually, while existing
group feature selection methods cannot handle online processing. Motivated by
this, we formulate the online group feature selection problem, and propose a
novel selection approach for this problem. Our proposed approach consists of
two stages: online intra-group selection and online inter-group selection. In
the intra-group selection, we use spectral analysis to select discriminative
features in each group when it arrives. In the inter-group selection, we use
Lasso to select a globally optimal subset of features. This 2-stage procedure
continues until there are no more features to come or some predefined stopping
conditions are met. Extensive experiments conducted on benchmark and real-world
data sets demonstrate that our proposed approach outperforms other
state-of-the-art online feature selection methods.
| [
{
"version": "v1",
"created": "Fri, 18 Apr 2014 12:51:24 GMT"
},
{
"version": "v2",
"created": "Tue, 30 Sep 2014 03:29:44 GMT"
},
{
"version": "v3",
"created": "Thu, 23 Oct 2014 03:31:43 GMT"
}
] | 2014-10-24T00:00:00 | [
[
"Jing",
"Wang",
""
],
[
"Zhong-Qiu",
"Zhao",
""
],
[
"Xuegang",
"Hu",
""
],
[
"Yiu-ming",
"Cheung",
""
],
[
"Meng",
"Wang",
""
],
[
"Xindong",
"Wu",
""
]
] | TITLE: Online Group Feature Selection
ABSTRACT: Online feature selection with dynamic features has become an active research
area in recent years. However, in some real-world applications such as image
analysis and email spam filtering, features may arrive by groups. Existing
online feature selection methods evaluate features individually, while existing
group feature selection methods cannot handle online processing. Motivated by
this, we formulate the online group feature selection problem, and propose a
novel selection approach for this problem. Our proposed approach consists of
two stages: online intra-group selection and online inter-group selection. In
the intra-group selection, we use spectral analysis to select discriminative
features in each group when it arrives. In the inter-group selection, we use
Lasso to select a globally optimal subset of features. This 2-stage procedure
continues until there are no more features to come or some predefined stopping
conditions are met. Extensive experiments conducted on benchmark and real-world
data sets demonstrate that our proposed approach outperforms other
state-of-the-art online feature selection methods.
|
1405.6879 | M\'arton Karsai | M\'arton Karsai, Gerardo I\~niguez, Kimmo Kaski, J\'anos Kert\'esz | Complex contagion process in spreading of online innovation | 27 pages, 11 figures, 2 tables | J. R. Soc. Interface 11, 101 (2014) | 10.1098/rsif.2014.0694 | null | physics.soc-ph cs.SI nlin.AO physics.comp-ph physics.data-an | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Diffusion of innovation can be interpreted as a social spreading phenomena
governed by the impact of media and social interactions. Although these
mechanisms have been identified by quantitative theories, their role and
relative importance are not entirely understood, since empirical verification
has so far been hindered by the lack of appropriate data. Here we analyse a
dataset recording the spreading dynamics of the world's largest Voice over
Internet Protocol service to empirically support the assumptions behind models
of social contagion. We show that the rate of spontaneous service adoption is
constant, the probability of adoption via social influence is linearly
proportional to the fraction of adopting neighbours, and the rate of service
termination is time-invariant and independent of the behaviour of peers. By
implementing the detected diffusion mechanisms into a dynamical agent-based
model, we are able to emulate the adoption dynamics of the service in several
countries worldwide. This approach enables us to make medium-term predictions
of service adoption and disclose dependencies between the dynamics of
innovation spreading and the socioeconomic development of a country.
| [
{
"version": "v1",
"created": "Tue, 27 May 2014 12:03:31 GMT"
},
{
"version": "v2",
"created": "Thu, 23 Oct 2014 10:10:21 GMT"
}
] | 2014-10-24T00:00:00 | [
[
"Karsai",
"Márton",
""
],
[
"Iñiguez",
"Gerardo",
""
],
[
"Kaski",
"Kimmo",
""
],
[
"Kertész",
"János",
""
]
] | TITLE: Complex contagion process in spreading of online innovation
ABSTRACT: Diffusion of innovation can be interpreted as a social spreading phenomena
governed by the impact of media and social interactions. Although these
mechanisms have been identified by quantitative theories, their role and
relative importance are not entirely understood, since empirical verification
has so far been hindered by the lack of appropriate data. Here we analyse a
dataset recording the spreading dynamics of the world's largest Voice over
Internet Protocol service to empirically support the assumptions behind models
of social contagion. We show that the rate of spontaneous service adoption is
constant, the probability of adoption via social influence is linearly
proportional to the fraction of adopting neighbours, and the rate of service
termination is time-invariant and independent of the behaviour of peers. By
implementing the detected diffusion mechanisms into a dynamical agent-based
model, we are able to emulate the adoption dynamics of the service in several
countries worldwide. This approach enables us to make medium-term predictions
of service adoption and disclose dependencies between the dynamics of
innovation spreading and the socioeconomic development of a country.
|
1409.5241 | Basura Fernando | Basura Fernando, Amaury Habrard, Marc Sebban and Tinne Tuytelaars | Subspace Alignment For Domain Adaptation | null | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this paper, we introduce a new domain adaptation (DA) algorithm where the
source and target domains are represented by subspaces spanned by eigenvectors.
Our method seeks a domain invariant feature space by learning a mapping
function which aligns the source subspace with the target one. We show that the
solution of the corresponding optimization problem can be obtained in a simple
closed form, leading to an extremely fast algorithm. We present two approaches
to determine the only hyper-parameter in our method corresponding to the size
of the subspaces. In the first approach we tune the size of subspaces using a
theoretical bound on the stability of the obtained result. In the second
approach, we use maximum likelihood estimation to determine the subspace size,
which is particularly useful for high dimensional data. Apart from PCA, we
propose a subspace creation method that outperform partial least squares (PLS)
and linear discriminant analysis (LDA) in domain adaptation. We test our method
on various datasets and show that, despite its intrinsic simplicity, it
outperforms state of the art DA methods.
| [
{
"version": "v1",
"created": "Thu, 18 Sep 2014 09:57:41 GMT"
},
{
"version": "v2",
"created": "Thu, 23 Oct 2014 08:40:06 GMT"
}
] | 2014-10-24T00:00:00 | [
[
"Fernando",
"Basura",
""
],
[
"Habrard",
"Amaury",
""
],
[
"Sebban",
"Marc",
""
],
[
"Tuytelaars",
"Tinne",
""
]
] | TITLE: Subspace Alignment For Domain Adaptation
ABSTRACT: In this paper, we introduce a new domain adaptation (DA) algorithm where the
source and target domains are represented by subspaces spanned by eigenvectors.
Our method seeks a domain invariant feature space by learning a mapping
function which aligns the source subspace with the target one. We show that the
solution of the corresponding optimization problem can be obtained in a simple
closed form, leading to an extremely fast algorithm. We present two approaches
to determine the only hyper-parameter in our method corresponding to the size
of the subspaces. In the first approach we tune the size of subspaces using a
theoretical bound on the stability of the obtained result. In the second
approach, we use maximum likelihood estimation to determine the subspace size,
which is particularly useful for high dimensional data. Apart from PCA, we
propose a subspace creation method that outperform partial least squares (PLS)
and linear discriminant analysis (LDA) in domain adaptation. We test our method
on various datasets and show that, despite its intrinsic simplicity, it
outperforms state of the art DA methods.
|
1410.4343 | Chiranjib Saha | Anupam Trivedi, Kunal Pal, Chiranjib Saha, Dipti Srinivasan | Enhanced Multiobjective Evolutionary Algorithm based on Decomposition
for Solving the Unit Commitment Problem | This paper has been withdrawn by the author due to 1. inconsistency
of vector representation in equation (12) and 2. algorithm 3 has not been
included in this text which makes the description incomplete | null | null | IEEE Transactions on Industrial Informatics - Manuscript ID
TII-14-0921 | cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | The unit commitment (UC) problem is a nonlinear, high-dimensional, highly
constrained, mixed-integer power system optimization problem and is generally
solved in the literature considering minimizing the system operation cost as
the only objective. However, due to increasing environmental concerns, the
recent attention has shifted to incorporating emission in the problem
formulation. In this paper, a multi-objective evolutionary algorithm based on
decomposition (MOEA/D) is proposed to solve the UC problem as a multi-objective
optimization problem considering minimizing cost and emission as the multiple
objec- tives. Since, UC problem is a mixed-integer optimization problem
consisting of binary UC variables and continuous power dispatch variables, a
novel hybridization strategy is proposed within the framework of MOEA/D such
that genetic algorithm (GA) evolves the binary variables while differential
evolution (DE) evolves the continuous variables. Further, a novel non-uniform
weight vector distribution strategy is proposed and a parallel island model
based on combination of MOEA/D with uniform and non-uniform weight vector
distribution strategy is implemented to enhance the performance of the
presented algorithm. Extensive case studies are presented on different test
systems and the effectiveness of the proposed hybridization strategy, the
non-uniform weight vector distribution strategy and parallel island model is
verified through stringent simulated results. Further, exhaustive benchmarking
against the algorithms proposed in the literature is presented to demonstrate
the superiority of the proposed algorithm in obtaining significantly better
converged and uniformly distributed trade-off solutions.
| [
{
"version": "v1",
"created": "Thu, 16 Oct 2014 09:17:51 GMT"
},
{
"version": "v2",
"created": "Thu, 23 Oct 2014 18:10:19 GMT"
}
] | 2014-10-24T00:00:00 | [
[
"Trivedi",
"Anupam",
""
],
[
"Pal",
"Kunal",
""
],
[
"Saha",
"Chiranjib",
""
],
[
"Srinivasan",
"Dipti",
""
]
] | TITLE: Enhanced Multiobjective Evolutionary Algorithm based on Decomposition
for Solving the Unit Commitment Problem
ABSTRACT: The unit commitment (UC) problem is a nonlinear, high-dimensional, highly
constrained, mixed-integer power system optimization problem and is generally
solved in the literature considering minimizing the system operation cost as
the only objective. However, due to increasing environmental concerns, the
recent attention has shifted to incorporating emission in the problem
formulation. In this paper, a multi-objective evolutionary algorithm based on
decomposition (MOEA/D) is proposed to solve the UC problem as a multi-objective
optimization problem considering minimizing cost and emission as the multiple
objec- tives. Since, UC problem is a mixed-integer optimization problem
consisting of binary UC variables and continuous power dispatch variables, a
novel hybridization strategy is proposed within the framework of MOEA/D such
that genetic algorithm (GA) evolves the binary variables while differential
evolution (DE) evolves the continuous variables. Further, a novel non-uniform
weight vector distribution strategy is proposed and a parallel island model
based on combination of MOEA/D with uniform and non-uniform weight vector
distribution strategy is implemented to enhance the performance of the
presented algorithm. Extensive case studies are presented on different test
systems and the effectiveness of the proposed hybridization strategy, the
non-uniform weight vector distribution strategy and parallel island model is
verified through stringent simulated results. Further, exhaustive benchmarking
against the algorithms proposed in the literature is presented to demonstrate
the superiority of the proposed algorithm in obtaining significantly better
converged and uniformly distributed trade-off solutions.
|
1311.2524 | Ross Girshick | Ross Girshick, Jeff Donahue, Trevor Darrell, Jitendra Malik | Rich feature hierarchies for accurate object detection and semantic
segmentation | Extended version of our CVPR 2014 paper; latest update (v5) includes
results using deeper networks (see Appendix G. Changelog) | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Object detection performance, as measured on the canonical PASCAL VOC
dataset, has plateaued in the last few years. The best-performing methods are
complex ensemble systems that typically combine multiple low-level image
features with high-level context. In this paper, we propose a simple and
scalable detection algorithm that improves mean average precision (mAP) by more
than 30% relative to the previous best result on VOC 2012---achieving a mAP of
53.3%. Our approach combines two key insights: (1) one can apply high-capacity
convolutional neural networks (CNNs) to bottom-up region proposals in order to
localize and segment objects and (2) when labeled training data is scarce,
supervised pre-training for an auxiliary task, followed by domain-specific
fine-tuning, yields a significant performance boost. Since we combine region
proposals with CNNs, we call our method R-CNN: Regions with CNN features. We
also compare R-CNN to OverFeat, a recently proposed sliding-window detector
based on a similar CNN architecture. We find that R-CNN outperforms OverFeat by
a large margin on the 200-class ILSVRC2013 detection dataset. Source code for
the complete system is available at http://www.cs.berkeley.edu/~rbg/rcnn.
| [
{
"version": "v1",
"created": "Mon, 11 Nov 2013 18:43:49 GMT"
},
{
"version": "v2",
"created": "Tue, 15 Apr 2014 01:44:31 GMT"
},
{
"version": "v3",
"created": "Wed, 7 May 2014 17:09:23 GMT"
},
{
"version": "v4",
"created": "Mon, 9 Jun 2014 22:07:33 GMT"
},
{
"version": "v5",
"created": "Wed, 22 Oct 2014 17:23:20 GMT"
}
] | 2014-10-23T00:00:00 | [
[
"Girshick",
"Ross",
""
],
[
"Donahue",
"Jeff",
""
],
[
"Darrell",
"Trevor",
""
],
[
"Malik",
"Jitendra",
""
]
] | TITLE: Rich feature hierarchies for accurate object detection and semantic
segmentation
ABSTRACT: Object detection performance, as measured on the canonical PASCAL VOC
dataset, has plateaued in the last few years. The best-performing methods are
complex ensemble systems that typically combine multiple low-level image
features with high-level context. In this paper, we propose a simple and
scalable detection algorithm that improves mean average precision (mAP) by more
than 30% relative to the previous best result on VOC 2012---achieving a mAP of
53.3%. Our approach combines two key insights: (1) one can apply high-capacity
convolutional neural networks (CNNs) to bottom-up region proposals in order to
localize and segment objects and (2) when labeled training data is scarce,
supervised pre-training for an auxiliary task, followed by domain-specific
fine-tuning, yields a significant performance boost. Since we combine region
proposals with CNNs, we call our method R-CNN: Regions with CNN features. We
also compare R-CNN to OverFeat, a recently proposed sliding-window detector
based on a similar CNN architecture. We find that R-CNN outperforms OverFeat by
a large margin on the 200-class ILSVRC2013 detection dataset. Source code for
the complete system is available at http://www.cs.berkeley.edu/~rbg/rcnn.
|
1408.5389 | Zhensong Qian | Zhensong Qian, Oliver Schulte and Yan Sun | Computing Multi-Relational Sufficient Statistics for Large Databases | 11pages, 8 figures, 8 tables, CIKM'14,November 3--7, 2014, Shanghai,
China | null | 10.1145/2661829.2662010 | null | cs.LG cs.DB | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Databases contain information about which relationships do and do not hold
among entities. To make this information accessible for statistical analysis
requires computing sufficient statistics that combine information from
different database tables. Such statistics may involve any number of {\em
positive and negative} relationships. With a naive enumeration approach,
computing sufficient statistics for negative relationships is feasible only for
small databases. We solve this problem with a new dynamic programming algorithm
that performs a virtual join, where the requisite counts are computed without
materializing join tables. Contingency table algebra is a new extension of
relational algebra, that facilitates the efficient implementation of this
M\"obius virtual join operation. The M\"obius Join scales to large datasets
(over 1M tuples) with complex schemas. Empirical evaluation with seven
benchmark datasets showed that information about the presence and absence of
links can be exploited in feature selection, association rule mining, and
Bayesian network learning.
| [
{
"version": "v1",
"created": "Fri, 22 Aug 2014 19:12:19 GMT"
}
] | 2014-10-23T00:00:00 | [
[
"Qian",
"Zhensong",
""
],
[
"Schulte",
"Oliver",
""
],
[
"Sun",
"Yan",
""
]
] | TITLE: Computing Multi-Relational Sufficient Statistics for Large Databases
ABSTRACT: Databases contain information about which relationships do and do not hold
among entities. To make this information accessible for statistical analysis
requires computing sufficient statistics that combine information from
different database tables. Such statistics may involve any number of {\em
positive and negative} relationships. With a naive enumeration approach,
computing sufficient statistics for negative relationships is feasible only for
small databases. We solve this problem with a new dynamic programming algorithm
that performs a virtual join, where the requisite counts are computed without
materializing join tables. Contingency table algebra is a new extension of
relational algebra, that facilitates the efficient implementation of this
M\"obius virtual join operation. The M\"obius Join scales to large datasets
(over 1M tuples) with complex schemas. Empirical evaluation with seven
benchmark datasets showed that information about the presence and absence of
links can be exploited in feature selection, association rule mining, and
Bayesian network learning.
|
1410.6039 | Roberto Sebastiani | Roberto Sebastiani and Silvia Tomasi | Optimization Modulo Theories with Linear Rational Costs | Submitted on january 2014 to ACM Transactions on Computational Logic,
currently under revision. arXiv admin note: text overlap with arXiv:1202.1409 | null | null | null | cs.LO | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In the contexts of automated reasoning (AR) and formal verification (FV),
important decision problems are effectively encoded into Satisfiability Modulo
Theories (SMT). In the last decade efficient SMT solvers have been developed
for several theories of practical interest (e.g., linear arithmetic, arrays,
bit-vectors). Surprisingly, little work has been done to extend SMT to deal
with optimization problems; in particular, we are not aware of any previous
work on SMT solvers able to produce solutions which minimize cost functions
over arithmetical variables. This is unfortunate, since some problems of
interest require this functionality.
In the work described in this paper we start filling this gap. We present and
discuss two general procedures for leveraging SMT to handle the minimization of
linear rational cost functions, combining SMT with standard minimization
techniques. We have implemented the procedures within the MathSAT SMT solver.
Due to the absence of competitors in the AR, FV and SMT domains, we have
experimentally evaluated our implementation against state-of-the-art tools for
the domain of linear generalized disjunctive programming (LGDP), which is
closest in spirit to our domain, on sets of problems which have been previously
proposed as benchmarks for the latter tools. The results show that our tool is
very competitive with, and often outperforms, these tools on these problems,
clearly demonstrating the potential of the approach.
| [
{
"version": "v1",
"created": "Wed, 22 Oct 2014 13:39:31 GMT"
}
] | 2014-10-23T00:00:00 | [
[
"Sebastiani",
"Roberto",
""
],
[
"Tomasi",
"Silvia",
""
]
] | TITLE: Optimization Modulo Theories with Linear Rational Costs
ABSTRACT: In the contexts of automated reasoning (AR) and formal verification (FV),
important decision problems are effectively encoded into Satisfiability Modulo
Theories (SMT). In the last decade efficient SMT solvers have been developed
for several theories of practical interest (e.g., linear arithmetic, arrays,
bit-vectors). Surprisingly, little work has been done to extend SMT to deal
with optimization problems; in particular, we are not aware of any previous
work on SMT solvers able to produce solutions which minimize cost functions
over arithmetical variables. This is unfortunate, since some problems of
interest require this functionality.
In the work described in this paper we start filling this gap. We present and
discuss two general procedures for leveraging SMT to handle the minimization of
linear rational cost functions, combining SMT with standard minimization
techniques. We have implemented the procedures within the MathSAT SMT solver.
Due to the absence of competitors in the AR, FV and SMT domains, we have
experimentally evaluated our implementation against state-of-the-art tools for
the domain of linear generalized disjunctive programming (LGDP), which is
closest in spirit to our domain, on sets of problems which have been previously
proposed as benchmarks for the latter tools. The results show that our tool is
very competitive with, and often outperforms, these tools on these problems,
clearly demonstrating the potential of the approach.
|
1410.6126 | German Ros | German Ros and Jose Alvarez and Julio Guerrero | Motion Estimation via Robust Decomposition with Constrained Rank | Submitted to IEEE TIP | null | null | null | cs.CV | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we address the problem of outlier detection for robust motion
estimation by using modern sparse-low-rank decompositions, i.e., Robust
PCA-like methods, to impose global rank constraints. Robust decompositions have
shown to be good at splitting a corrupted matrix into an uncorrupted low-rank
matrix and a sparse matrix, containing outliers. However, this process only
works when matrices have relatively low rank with respect to their ambient
space, a property not met in motion estimation problems. As a solution, we
propose to exploit the partial information present in the decomposition to
decide which matches are outliers. We provide evidences showing that even when
it is not possible to recover an uncorrupted low-rank matrix, the resulting
information can be exploited for outlier detection. To this end we propose the
Robust Decomposition with Constrained Rank (RD-CR), a proximal gradient based
method that enforces the rank constraints inherent to motion estimation. We
also present a general framework to perform robust estimation for stereo Visual
Odometry, based on our RD-CR and a simple but effective compressed optimization
method that achieves high performance. Our evaluation on synthetic data and on
the KITTI dataset demonstrates the applicability of our approach in complex
scenarios and it yields state-of-the-art performance.
| [
{
"version": "v1",
"created": "Wed, 22 Oct 2014 18:15:27 GMT"
}
] | 2014-10-23T00:00:00 | [
[
"Ros",
"German",
""
],
[
"Alvarez",
"Jose",
""
],
[
"Guerrero",
"Julio",
""
]
] | TITLE: Motion Estimation via Robust Decomposition with Constrained Rank
ABSTRACT: In this work, we address the problem of outlier detection for robust motion
estimation by using modern sparse-low-rank decompositions, i.e., Robust
PCA-like methods, to impose global rank constraints. Robust decompositions have
shown to be good at splitting a corrupted matrix into an uncorrupted low-rank
matrix and a sparse matrix, containing outliers. However, this process only
works when matrices have relatively low rank with respect to their ambient
space, a property not met in motion estimation problems. As a solution, we
propose to exploit the partial information present in the decomposition to
decide which matches are outliers. We provide evidences showing that even when
it is not possible to recover an uncorrupted low-rank matrix, the resulting
information can be exploited for outlier detection. To this end we propose the
Robust Decomposition with Constrained Rank (RD-CR), a proximal gradient based
method that enforces the rank constraints inherent to motion estimation. We
also present a general framework to perform robust estimation for stereo Visual
Odometry, based on our RD-CR and a simple but effective compressed optimization
method that achieves high performance. Our evaluation on synthetic data and on
the KITTI dataset demonstrates the applicability of our approach in complex
scenarios and it yields state-of-the-art performance.
|
1406.6597 | Vincent Labatut | G\"unce Keziban Orman, Vincent Labatut, Marc Plantevit (LIRIS),
Jean-Fran\c{c}ois Boulicaut (LIRIS) | A Method for Characterizing Communities in Dynamic Attributed Complex
Networks | IEEE/ACM International Conference on Advances in Social Network
Analysis and Mining (ASONAM), P\'ekin : China (2014) | null | 10.1109/ASONAM.2014.6921629 | null | cs.SI physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Many methods have been proposed to detect communities, not only in plain, but
also in attributed, directed or even dynamic complex networks. In its simplest
form, a community structure takes the form of a partition of the node set. From
the modeling point of view, to be of some utility, this partition must then be
characterized relatively to the properties of the studied system. However, if
most of the existing works focus on defining methods for the detection of
communities, only very few try to tackle this interpretation problem. Moreover,
the existing approaches are limited either in the type of data they handle, or
by the nature of the results they output. In this work, we propose a method to
efficiently support such a characterization task. We first define a
sequence-based representation of networks, combining temporal information,
topological measures, and nodal attributes. We then describe how to identify
the most emerging sequential patterns of this dataset, and use them to
characterize the communities. We also show how to detect unusual behavior in a
community, and highlight outliers. Finally, as an illustration, we apply our
method to a network of scientific collaborations.
| [
{
"version": "v1",
"created": "Wed, 25 Jun 2014 14:54:52 GMT"
}
] | 2014-10-22T00:00:00 | [
[
"Orman",
"Günce Keziban",
"",
"LIRIS"
],
[
"Labatut",
"Vincent",
"",
"LIRIS"
],
[
"Plantevit",
"Marc",
"",
"LIRIS"
],
[
"Boulicaut",
"Jean-François",
"",
"LIRIS"
]
] | TITLE: A Method for Characterizing Communities in Dynamic Attributed Complex
Networks
ABSTRACT: Many methods have been proposed to detect communities, not only in plain, but
also in attributed, directed or even dynamic complex networks. In its simplest
form, a community structure takes the form of a partition of the node set. From
the modeling point of view, to be of some utility, this partition must then be
characterized relatively to the properties of the studied system. However, if
most of the existing works focus on defining methods for the detection of
communities, only very few try to tackle this interpretation problem. Moreover,
the existing approaches are limited either in the type of data they handle, or
by the nature of the results they output. In this work, we propose a method to
efficiently support such a characterization task. We first define a
sequence-based representation of networks, combining temporal information,
topological measures, and nodal attributes. We then describe how to identify
the most emerging sequential patterns of this dataset, and use them to
characterize the communities. We also show how to detect unusual behavior in a
community, and highlight outliers. Finally, as an illustration, we apply our
method to a network of scientific collaborations.
|
1410.5467 | Josef Urban | Cezary Kaliszyk, Lionel Mamane, Josef Urban | Machine Learning of Coq Proof Guidance: First Experiments | null | null | null | null | cs.LO cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We report the results of the first experiments with learning proof
dependencies from the formalizations done with the Coq system. We explain the
process of obtaining the dependencies from the Coq proofs, the characterization
of formulas that is used for the learning, and the evaluation method. Various
machine learning methods are compared on a dataset of 5021 toplevel Coq proofs
coming from the CoRN repository. The best resulting method covers on average
75% of the needed proof dependencies among the first 100 predictions, which is
a comparable performance of such initial experiments on other large-theory
corpora.
| [
{
"version": "v1",
"created": "Mon, 20 Oct 2014 21:16:52 GMT"
}
] | 2014-10-22T00:00:00 | [
[
"Kaliszyk",
"Cezary",
""
],
[
"Mamane",
"Lionel",
""
],
[
"Urban",
"Josef",
""
]
] | TITLE: Machine Learning of Coq Proof Guidance: First Experiments
ABSTRACT: We report the results of the first experiments with learning proof
dependencies from the formalizations done with the Coq system. We explain the
process of obtaining the dependencies from the Coq proofs, the characterization
of formulas that is used for the learning, and the evaluation method. Various
machine learning methods are compared on a dataset of 5021 toplevel Coq proofs
coming from the CoRN repository. The best resulting method covers on average
75% of the needed proof dependencies among the first 100 predictions, which is
a comparable performance of such initial experiments on other large-theory
corpora.
|
1410.5476 | Josef Urban | Cezary Kaliszyk, Josef Urban, Jiri Vyskocil | Certified Connection Tableaux Proofs for HOL Light and TPTP | null | null | null | null | cs.LO cs.AI | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In the recent years, the Metis prover based on ordered paramodulation and
model elimination has replaced the earlier built-in methods for general-purpose
proof automation in HOL4 and Isabelle/HOL. In the annual CASC competition, the
leanCoP system based on connection tableaux has however performed better than
Metis. In this paper we show how the leanCoP's core algorithm can be
implemented inside HOLLight. leanCoP's flagship feature, namely its
minimalistic core, results in a very simple proof system. This plays a crucial
role in extending the MESON proof reconstruction mechanism to connection
tableaux proofs, providing an implementation of leanCoP that certifies its
proofs. We discuss the differences between our direct implementation using an
explicit Prolog stack, to the continuation passing implementation of MESON
present in HOLLight and compare their performance on all core HOLLight goals.
The resulting prover can be also used as a general purpose TPTP prover. We
compare its performance against the resolution based Metis on TPTP and other
interesting datasets.
| [
{
"version": "v1",
"created": "Mon, 20 Oct 2014 21:36:47 GMT"
}
] | 2014-10-22T00:00:00 | [
[
"Kaliszyk",
"Cezary",
""
],
[
"Urban",
"Josef",
""
],
[
"Vyskocil",
"Jiri",
""
]
] | TITLE: Certified Connection Tableaux Proofs for HOL Light and TPTP
ABSTRACT: In the recent years, the Metis prover based on ordered paramodulation and
model elimination has replaced the earlier built-in methods for general-purpose
proof automation in HOL4 and Isabelle/HOL. In the annual CASC competition, the
leanCoP system based on connection tableaux has however performed better than
Metis. In this paper we show how the leanCoP's core algorithm can be
implemented inside HOLLight. leanCoP's flagship feature, namely its
minimalistic core, results in a very simple proof system. This plays a crucial
role in extending the MESON proof reconstruction mechanism to connection
tableaux proofs, providing an implementation of leanCoP that certifies its
proofs. We discuss the differences between our direct implementation using an
explicit Prolog stack, to the continuation passing implementation of MESON
present in HOLLight and compare their performance on all core HOLLight goals.
The resulting prover can be also used as a general purpose TPTP prover. We
compare its performance against the resolution based Metis on TPTP and other
interesting datasets.
|
1410.5652 | Andr\'as Kir\'aly Ph.D. | Tam\'as Varga, Andr\'as Kir\'aly, J\'anos Abonyi | Improvement of PSO algorithm by memory based gradient search -
application in inventory management | book chapter, 20 pages, 7 figures, 2 tables | Swarm Intelligence and Bio-Inspired Computation: Theory and
Applications, 1st Edition, pages 403-422. Elsevier, Oxford, 2013 | null | null | cs.NE | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Advanced inventory management in complex supply chains requires effective and
robust nonlinear optimization due to the stochastic nature of supply and demand
variations. Application of estimated gradients can boost up the convergence of
Particle Swarm Optimization (PSO) algorithm but classical gradient calculation
cannot be applied to stochastic and uncertain systems. In these situations
Monte-Carlo (MC) simulation can be applied to determine the gradient. We
developed a memory based algorithm where instead of generating and evaluating
new simulated samples the stored and shared former function evaluations of the
particles are sampled to estimate the gradients by local weighted least squares
regression. The performance of the resulted regional gradient-based PSO is
verified by several benchmark problems and in a complex application example
where optimal reorder points of a supply chain are determined.
| [
{
"version": "v1",
"created": "Tue, 21 Oct 2014 13:40:23 GMT"
}
] | 2014-10-22T00:00:00 | [
[
"Varga",
"Tamás",
""
],
[
"Király",
"András",
""
],
[
"Abonyi",
"János",
""
]
] | TITLE: Improvement of PSO algorithm by memory based gradient search -
application in inventory management
ABSTRACT: Advanced inventory management in complex supply chains requires effective and
robust nonlinear optimization due to the stochastic nature of supply and demand
variations. Application of estimated gradients can boost up the convergence of
Particle Swarm Optimization (PSO) algorithm but classical gradient calculation
cannot be applied to stochastic and uncertain systems. In these situations
Monte-Carlo (MC) simulation can be applied to determine the gradient. We
developed a memory based algorithm where instead of generating and evaluating
new simulated samples the stored and shared former function evaluations of the
particles are sampled to estimate the gradients by local weighted least squares
regression. The performance of the resulted regional gradient-based PSO is
verified by several benchmark problems and in a complex application example
where optimal reorder points of a supply chain are determined.
|
1410.5684 | Saahil Ognawala | Saahil Ognawala and Justin Bayer | Regularizing Recurrent Networks - On Injected Noise and Norm-based
Methods | null | null | null | null | stat.ML cs.LG | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Advancements in parallel processing have lead to a surge in multilayer
perceptrons' (MLP) applications and deep learning in the past decades.
Recurrent Neural Networks (RNNs) give additional representational power to
feedforward MLPs by providing a way to treat sequential data. However, RNNs are
hard to train using conventional error backpropagation methods because of the
difficulty in relating inputs over many time-steps. Regularization approaches
from MLP sphere, like dropout and noisy weight training, have been
insufficiently applied and tested on simple RNNs. Moreover, solutions have been
proposed to improve convergence in RNNs but not enough to improve the long term
dependency remembering capabilities thereof.
In this study, we aim to empirically evaluate the remembering and
generalization ability of RNNs on polyphonic musical datasets. The models are
trained with injected noise, random dropout, norm-based regularizers and their
respective performances compared to well-initialized plain RNNs and advanced
regularization methods like fast-dropout. We conclude with evidence that
training with noise does not improve performance as conjectured by a few works
in RNN optimization before ours.
| [
{
"version": "v1",
"created": "Tue, 21 Oct 2014 14:36:26 GMT"
}
] | 2014-10-22T00:00:00 | [
[
"Ognawala",
"Saahil",
""
],
[
"Bayer",
"Justin",
""
]
] | TITLE: Regularizing Recurrent Networks - On Injected Noise and Norm-based
Methods
ABSTRACT: Advancements in parallel processing have lead to a surge in multilayer
perceptrons' (MLP) applications and deep learning in the past decades.
Recurrent Neural Networks (RNNs) give additional representational power to
feedforward MLPs by providing a way to treat sequential data. However, RNNs are
hard to train using conventional error backpropagation methods because of the
difficulty in relating inputs over many time-steps. Regularization approaches
from MLP sphere, like dropout and noisy weight training, have been
insufficiently applied and tested on simple RNNs. Moreover, solutions have been
proposed to improve convergence in RNNs but not enough to improve the long term
dependency remembering capabilities thereof.
In this study, we aim to empirically evaluate the remembering and
generalization ability of RNNs on polyphonic musical datasets. The models are
trained with injected noise, random dropout, norm-based regularizers and their
respective performances compared to well-initialized plain RNNs and advanced
regularization methods like fast-dropout. We conclude with evidence that
training with noise does not improve performance as conjectured by a few works
in RNN optimization before ours.
|
1302.6636 | Tamara Kolda | Tamara G. Kolda and Ali Pinar and Todd Plantenga and C. Seshadhri | A Scalable Generative Graph Model with Community Structure | null | SIAM Journal on Scientific Computing, Vol. 36, No. 5, pp.
C424-C452, September 2014 | 10.1137/130914218 | null | cs.SI physics.soc-ph | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Network data is ubiquitous and growing, yet we lack realistic generative
network models that can be calibrated to match real-world data. The recently
proposed Block Two-Level Erdss-Renyi (BTER) model can be tuned to capture two
fundamental properties: degree distribution and clustering coefficients. The
latter is particularly important for reproducing graphs with community
structure, such as social networks. In this paper, we compare BTER to other
scalable models and show that it gives a better fit to real data. We provide a
scalable implementation that requires only O(d_max) storage where d_max is the
maximum number of neighbors for a single node. The generator is trivially
parallelizable, and we show results for a Hadoop MapReduce implementation for a
modeling a real-world web graph with over 4.6 billion edges. We propose that
the BTER model can be used as a graph generator for benchmarking purposes and
provide idealized degree distributions and clustering coefficient profiles that
can be tuned for user specifications.
| [
{
"version": "v1",
"created": "Wed, 27 Feb 2013 00:53:12 GMT"
},
{
"version": "v2",
"created": "Fri, 22 Mar 2013 21:11:37 GMT"
},
{
"version": "v3",
"created": "Tue, 10 Dec 2013 01:22:52 GMT"
}
] | 2014-10-21T00:00:00 | [
[
"Kolda",
"Tamara G.",
""
],
[
"Pinar",
"Ali",
""
],
[
"Plantenga",
"Todd",
""
],
[
"Seshadhri",
"C.",
""
]
] | TITLE: A Scalable Generative Graph Model with Community Structure
ABSTRACT: Network data is ubiquitous and growing, yet we lack realistic generative
network models that can be calibrated to match real-world data. The recently
proposed Block Two-Level Erdss-Renyi (BTER) model can be tuned to capture two
fundamental properties: degree distribution and clustering coefficients. The
latter is particularly important for reproducing graphs with community
structure, such as social networks. In this paper, we compare BTER to other
scalable models and show that it gives a better fit to real data. We provide a
scalable implementation that requires only O(d_max) storage where d_max is the
maximum number of neighbors for a single node. The generator is trivially
parallelizable, and we show results for a Hadoop MapReduce implementation for a
modeling a real-world web graph with over 4.6 billion edges. We propose that
the BTER model can be used as a graph generator for benchmarking purposes and
provide idealized degree distributions and clustering coefficient profiles that
can be tuned for user specifications.
|
1312.3968 | Philip Schniter | Mark Borgerding, Philip Schniter, and Sundeep Rangan | Generalized Approximate Message Passing for Cosparse Analysis
Compressive Sensing | null | null | null | null | cs.IT math.IT | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In cosparse analysis compressive sensing (CS), one seeks to estimate a
non-sparse signal vector from noisy sub-Nyquist linear measurements by
exploiting the knowledge that a given linear transform of the signal is
cosparse, i.e., has sufficiently many zeros. We propose a novel approach to
cosparse analysis CS based on the generalized approximate message passing
(GAMP) algorithm. Unlike other AMP-based approaches to this problem, ours works
with a wide range of analysis operators and regularizers. In addition, we
propose a novel $\ell_0$-like soft-thresholder based on MMSE denoising for a
spike-and-slab distribution with an infinite-variance slab. Numerical
demonstrations on synthetic and practical datasets demonstrate advantages over
existing AMP-based, greedy, and reweighted-$\ell_1$ approaches.
| [
{
"version": "v1",
"created": "Fri, 13 Dec 2013 21:51:20 GMT"
},
{
"version": "v2",
"created": "Sun, 19 Oct 2014 19:12:54 GMT"
}
] | 2014-10-21T00:00:00 | [
[
"Borgerding",
"Mark",
""
],
[
"Schniter",
"Philip",
""
],
[
"Rangan",
"Sundeep",
""
]
] | TITLE: Generalized Approximate Message Passing for Cosparse Analysis
Compressive Sensing
ABSTRACT: In cosparse analysis compressive sensing (CS), one seeks to estimate a
non-sparse signal vector from noisy sub-Nyquist linear measurements by
exploiting the knowledge that a given linear transform of the signal is
cosparse, i.e., has sufficiently many zeros. We propose a novel approach to
cosparse analysis CS based on the generalized approximate message passing
(GAMP) algorithm. Unlike other AMP-based approaches to this problem, ours works
with a wide range of analysis operators and regularizers. In addition, we
propose a novel $\ell_0$-like soft-thresholder based on MMSE denoising for a
spike-and-slab distribution with an infinite-variance slab. Numerical
demonstrations on synthetic and practical datasets demonstrate advantages over
existing AMP-based, greedy, and reweighted-$\ell_1$ approaches.
|
1410.4966 | Shay Cohen | Chiraag Lala and Shay B. Cohen | The Visualization of Change in Word Meaning over Time using Temporal
Word Embeddings | null | null | null | null | cs.CL | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | We describe a visualization tool that can be used to view the change in
meaning of words over time. The tool makes use of existing (static) word
embedding datasets together with a timestamped $n$-gram corpus to create {\em
temporal} word embeddings.
| [
{
"version": "v1",
"created": "Sat, 18 Oct 2014 14:53:19 GMT"
}
] | 2014-10-21T00:00:00 | [
[
"Lala",
"Chiraag",
""
],
[
"Cohen",
"Shay B.",
""
]
] | TITLE: The Visualization of Change in Word Meaning over Time using Temporal
Word Embeddings
ABSTRACT: We describe a visualization tool that can be used to view the change in
meaning of words over time. The tool makes use of existing (static) word
embedding datasets together with a timestamped $n$-gram corpus to create {\em
temporal} word embeddings.
|
1410.4984 | Zhenwen Dai | Zhenwen Dai, Andreas Damianou, James Hensman, Neil Lawrence | Gaussian Process Models with Parallelization and GPU acceleration | null | null | null | null | cs.DC cs.LG stat.ML | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | In this work, we present an extension of Gaussian process (GP) models with
sophisticated parallelization and GPU acceleration. The parallelization scheme
arises naturally from the modular computational structure w.r.t. datapoints in
the sparse Gaussian process formulation. Additionally, the computational
bottleneck is implemented with GPU acceleration for further speed up. Combining
both techniques allows applying Gaussian process models to millions of
datapoints. The efficiency of our algorithm is demonstrated with a synthetic
dataset. Its source code has been integrated into our popular software library
GPy.
| [
{
"version": "v1",
"created": "Sat, 18 Oct 2014 18:12:57 GMT"
}
] | 2014-10-21T00:00:00 | [
[
"Dai",
"Zhenwen",
""
],
[
"Damianou",
"Andreas",
""
],
[
"Hensman",
"James",
""
],
[
"Lawrence",
"Neil",
""
]
] | TITLE: Gaussian Process Models with Parallelization and GPU acceleration
ABSTRACT: In this work, we present an extension of Gaussian process (GP) models with
sophisticated parallelization and GPU acceleration. The parallelization scheme
arises naturally from the modular computational structure w.r.t. datapoints in
the sparse Gaussian process formulation. Additionally, the computational
bottleneck is implemented with GPU acceleration for further speed up. Combining
both techniques allows applying Gaussian process models to millions of
datapoints. The efficiency of our algorithm is demonstrated with a synthetic
dataset. Its source code has been integrated into our popular software library
GPy.
|
1403.1349 | Sam Anzaroot | Sam Anzaroot, Alexandre Passos, David Belanger, Andrew McCallum | Learning Soft Linear Constraints with Application to Citation Field
Extraction | appears in Proc. the 52nd Annual Meeting of the Association for
Computational Linguistics (ACL2014) | null | null | null | cs.CL cs.DL cs.IR | http://arxiv.org/licenses/nonexclusive-distrib/1.0/ | Accurately segmenting a citation string into fields for authors, titles, etc.
is a challenging task because the output typically obeys various global
constraints. Previous work has shown that modeling soft constraints, where the
model is encouraged, but not require to obey the constraints, can substantially
improve segmentation performance. On the other hand, for imposing hard
constraints, dual decomposition is a popular technique for efficient prediction
given existing algorithms for unconstrained inference. We extend the technique
to perform prediction subject to soft constraints. Moreover, with a technique
for performing inference given soft constraints, it is easy to automatically
generate large families of constraints and learn their costs with a simple
convex optimization problem during training. This allows us to obtain
substantial gains in accuracy on a new, challenging citation extraction
dataset.
| [
{
"version": "v1",
"created": "Thu, 6 Mar 2014 05:24:02 GMT"
},
{
"version": "v2",
"created": "Fri, 17 Oct 2014 13:27:02 GMT"
}
] | 2014-10-20T00:00:00 | [
[
"Anzaroot",
"Sam",
""
],
[
"Passos",
"Alexandre",
""
],
[
"Belanger",
"David",
""
],
[
"McCallum",
"Andrew",
""
]
] | TITLE: Learning Soft Linear Constraints with Application to Citation Field
Extraction
ABSTRACT: Accurately segmenting a citation string into fields for authors, titles, etc.
is a challenging task because the output typically obeys various global
constraints. Previous work has shown that modeling soft constraints, where the
model is encouraged, but not require to obey the constraints, can substantially
improve segmentation performance. On the other hand, for imposing hard
constraints, dual decomposition is a popular technique for efficient prediction
given existing algorithms for unconstrained inference. We extend the technique
to perform prediction subject to soft constraints. Moreover, with a technique
for performing inference given soft constraints, it is easy to automatically
generate large families of constraints and learn their costs with a simple
convex optimization problem during training. This allows us to obtain
substantial gains in accuracy on a new, challenging citation extraction
dataset.
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.