id
stringlengths
9
16
title
stringlengths
4
278
categories
stringlengths
5
104
abstract
stringlengths
6
4.09k
1004.4758
A Design of Paraunitary Polyphase Matrices of Rational Filter Banks Based on (P,Q) Shift-Invariant Systems
cs.IT math.IT
In this paper we present a method to design paraunitary polyphase matrices of critically sampled rational filter banks. The method is based on (P,Q) shift-invariant systems, and so any kind of rational splitting of the frequency spectrum can be achieved using this method. Ideal (P,Q) shift-invariant system with smallest P and Q that map of a band of input spectrum to the output spectrum are obtained. A new set of filters is obtained that characterize a (P,Q) shift-invariant system. Ideal frequency spectrum of these filters are obtained using ideal $(P,Q)$ shift-invariant systems. Actual paraunitary polyphase matrices are then obtained by minimizing the stopband energies of these filters against the parameters of the paraunitary polyphase matrices.
1004.4793
Logical methods of object recognition on satellite images using spatial constraints
cs.CV
A logical approach to object recognition on image is proposed. The main idea of the approach is to perform the object recognition as a logical inference on a set of rules describing an object shape.
1004.4801
Ontology-based inference for causal explanation
cs.AI
We define an inference system to capture explanations based on causal statements, using an ontology in the form of an IS-A hierarchy. We first introduce a simple logical language which makes it possible to express that a fact causes another fact and that a fact explains another fact. We present a set of formal inference patterns from causal statements to explanation statements. We introduce an elementary ontology which gives greater expressiveness to the system while staying close to propositional reasoning. We provide an inference system that captures the patterns discussed, firstly in a purely propositional framework, then in a datalog (limited predicate) framework.
1004.4815
Universal A Posteriori Metrics Game
cs.IT math.IT math.PR
Over binary input channels, uniform distribution is a universal prior, in the sense that it allows to maximize the worst case mutual information over all binary input channels, ensuring at least 94.2% of the capacity. In this paper, we address a similar question, but with respect to a universal generalized linear decoder. We look for the best collection of finitely many a posteriori metrics, to maximize the worst case mismatched mutual information achieved by decoding with these metrics (instead of an optimal decoder such as the Maximum Likelihood (ML) tuned to the true channel). It is shown that for binary input and output channels, two metrics suffice to actually achieve the same performance as an optimal decoder. In particular, this implies that there exist a decoder which is generalized linear and achieves at least 94.2% of the compound capacity on any compound set, without the knowledge of the underlying set.
1004.4824
Growth and structure of Slovenia's scientific collaboration network
physics.soc-ph cond-mat.stat-mech cs.DB
We study the evolution of Slovenia's scientific collaboration network from 1960 till present with a yearly resolution. For each year the network was constructed from publication records of Slovene scientists, whereby two were connected if, up to the given year inclusive, they have coauthored at least one paper together. Starting with no more than 30 scientists with an average of 1.5 collaborators in the year 1960, the network to date consists of 7380 individuals that, on average, have 10.7 collaborators. We show that, in spite of the broad myriad of research fields covered, the networks form "small worlds" and that indeed the average path between any pair of scientists scales logarithmically with size after the largest component becomes large enough. Moreover, we show that the network growth is governed by near-liner preferential attachment, giving rise to a log-normal distribution of collaborators per author, and that the average starting year is roughly inversely proportional to the number of collaborators eventually acquired. Understandably, not all that became active early have till now gathered many collaborators. We also give results for the clustering coefficient and the diameter of the network over time, and compare our conclusions with those reported previously.
1004.4826
Impact of Channel Asymmetry on Base Station Cooperative Transmission with Limited Feedback
cs.IT math.IT
Base station (BS) cooperative transmission, also known as coordinated multi-point transmission (CoMP), is an effective way to avoid inter-cell interference in universal frequency reuse cellular systems. To gain the promised benefit, however, huge feedback overhead is in demand to gather the channel information. In this paper, we analyze the impact of channel asymmetry, which is inherent in CoMP systems, on downlink BS cooperative transmission with limited feedback. We analyze the per-user rate loss of a multi-user CoMP system led by quantization. Per-cell quantization of multicell channels is considered, which quantizes the local channel and cross channel separately and is more feasible in practice. From both the analytical and simulation results, we provide a whole picture on various critical factors that lead to the performance loss. Specifically, we show that the per user rate loss led by limited feedback depends on the location of its paired users, except for relying on its own signal to noise ratio and the quantization errors as in single cell multi-user multiple antenna systems. This implies that the quantization accuracy required for local and cross channel of each user depends on the locations of its own as well as its paired users.
1004.4830
Performance Evaluation of SCM-WDM System Using Different Linecoding
cs.OH cs.IT math.IT
This paper investigates the theoretical performance analysis for a subcarrier multiplexed (SCM) wavelength division multiplexing (WDM) optical transmission system in presence of optical beat interference (OBI) which occurs during the photo detection process. We have presented a comparison for improving the performance of SCM-WDM system in presence of OBI. Non-return-to zero (NRZ), Manchester and Miller code (MC) line coding are used for performance investigation of SCM-WDM system. A suitable signal bandwidth is selected and 200 KHz is considered as channel bandwidth. Power spectrum of signal and cross component for those line coding are analyzed. Comparison results are evaluated in terms of signal to OBI ratio for the three linecoding schemes which is called signal to interference ratio (SIR). It is found that there is a significant increase in the SIR by employing Miller code compared to NRZ and Manchester for the same data rate. For example, for a number of subcarriers of 10, the achievable SIR is about -24 dB for Miller coded system compared to -46 dB for NRZ coded system and -49 dB for Manchester coded system. The results are found to be satisfactorily agreed with the expected results.
1004.4848
Punctuation effects in English and Esperanto texts
cs.CL physics.data-an
A statistical physics study of punctuation effects on sentence lengths is presented for written texts: {\it Alice in wonderland} and {\it Through a looking glass}. The translation of the first text into esperanto is also considered as a test for the role of punctuation in defining a style, and for contrasting natural and artificial, but written, languages. Several log-log plots of the sentence length-rank relationship are presented for the major punctuation marks. Different power laws are observed with characteristic exponents. The exponent can take a value much less than unity ($ca.$ 0.50 or 0.30) depending on how a sentence is defined. The texts are also mapped into time series based on the word frequencies. The quantitative differences between the original and translated texts are very minutes, at the exponent level. It is argued that sentences seem to be more reliable than word distributions in discussing an author style.
1004.4864
Polynomial Learning of Distribution Families
cs.LG cs.DS
The question of polynomial learnability of probability distributions, particularly Gaussian mixture distributions, has recently received significant attention in theoretical computer science and machine learning. However, despite major progress, the general question of polynomial learnability of Gaussian mixture distributions still remained open. The current work resolves the question of polynomial learnability for Gaussian mixtures in high dimension with an arbitrary fixed number of components. The result on learning Gaussian mixtures relies on an analysis of distributions belonging to what we call "polynomial families" in low dimension. These families are characterized by their moments being polynomial in parameters and include almost all common probability distributions as well as their mixtures and products. Using tools from real algebraic geometry, we show that parameters of any distribution belonging to such a family can be learned in polynomial time and using a polynomial number of sample points. The result on learning polynomial families is quite general and is of independent interest. To estimate parameters of a Gaussian mixture distribution in high dimensions, we provide a deterministic algorithm for dimensionality reduction. This allows us to reduce learning a high-dimensional mixture to a polynomial number of parameter estimations in low dimension. Combining this reduction with the results on polynomial families yields our result on learning arbitrary Gaussian mixtures in high dimensions.
1004.4880
ECME Thresholding Methods for Sparse Signal Reconstruction
cs.IT math.IT
We propose a probabilistic framework for interpreting and developing hard thresholding sparse signal reconstruction methods and present several new algorithms based on this framework. The measurements follow an underdetermined linear model, where the regression-coefficient vector is the sum of an unknown deterministic sparse signal component and a zero-mean white Gaussian component with an unknown variance. We first derive an expectation-conditional maximization either (ECME) iteration that guarantees convergence to a local maximum of the likelihood function of the unknown parameters for a given signal sparsity level. To analyze the reconstruction accuracy, we introduce the minimum sparse subspace quotient (SSQ), a more flexible measure of the sampling operator than the well-established restricted isometry property (RIP). We prove that, if the minimum SSQ is sufficiently large, ECME achieves perfect or near-optimal recovery of sparse or approximately sparse signals, respectively. We also propose a double overrelaxation (DORE) thresholding scheme for accelerating the ECME iteration. If the signal sparsity level is unknown, we introduce an unconstrained sparsity selection (USS) criterion for its selection and show that, under certain conditions, applying this criterion is equivalent to finding the sparsest solution of the underlying underdetermined linear system. Finally, we present our automatic double overrelaxation (ADORE) thresholding method that utilizes the USS criterion to select the signal sparsity level. We apply the proposed schemes to reconstruct sparse and approximately sparse signals from tomographic projections and compressive samples.
1004.4882
Properties of Codes in the Johnson Scheme
cs.IT math.IT
Codes which attain the sphere packing bound are called perfect codes. The most important metrics in coding theory on which perfect codes are defined are the Hamming metric and the Johnson metric. While for the Hamming metric all perfect codes over finite fields are known, in the Johnson metric it was conjectured by Delsarte in 1970's that there are no nontrivial perfect codes. The general nonexistence proof still remains the open problem. In this work we examine constant weight codes as well as doubly constant weight codes, and reduce the range of parameters in which perfect codes may exist in both cases. We start with the constant weight codes. We introduce an improvement of Roos' bound for one-perfect codes, and present some new divisibility conditions, which are based on the connection between perfect codes in Johnson graph J(n,w) and block designs. Next, we consider binomial moments for perfect codes. We show which parameters can be excluded for one-perfect codes. We examine two-perfect codes in J(2w,w) and present necessary conditions for existence of such codes. We prove that there are no two-perfect codes in J(2w,w) with length less then 2.5*10^{15}. Next we examine perfect doubly constant weight codes. We present a family of parameters for codes whose size of sphere divides the size of whole space. We then prove a bound on length of such codes, similarly to Roos' bound for perfect codes in Johnson graph. Finally we describe Steiner systems and doubly Steiner systems, which are strongly connected with the constant weight and doubly constant weight codes respectively. We provide an anticode-based proof of a bound on length of Steiner system, prove that doubly Steiner system is a diameter perfect code and present a bound on length of doubly Steiner system.
1004.4917
On the Capacity of Compound State-Dependent Channels with States Known at the Transmitter
cs.IT math.IT math.PR
This paper investigates the capacity of compound state-dependent channels with non-causal state information available at only the transmitter. A new lower bound on the capacity of this class of channels is derived. This bound is shown to be tight for the special case of compound channels with stochastic degraded components, yielding the full characterization of the capacity. Specific results are derived for the compound Gaussian Dirty-Paper (GDP) channel. This model consists of an additive white Gaussian noise (AWGN) channel corrupted by an additive Gaussian interfering signal, known at the transmitter only, where the input and the state signals are affected by fading coefficients whose realizations are unknown at the transmitter. Our bounds are shown to be tight for specific cases. Applications of these results arise in a variety of wireless scenarios as multicast channels, cognitive radio and problems with interference cancellation.
1004.4944
Outer Bounds for the Interference Channel with a Cognitive Relay
cs.IT math.IT
In this paper, we first present an outer bound for a general interference channel with a cognitive relay, i.e., a relay that has non-causal knowledge of both independent messages transmitted in the interference channel. This outer bound reduces to the capacity region of the deterministic broadcast channel and of the deterministic cognitive interference channel through nulling of certain channel inputs. It does not, however, reduce to that of certain deterministic interference channels for which capacity is known. As such, we subsequently tighten the bound for channels whose outputs satisfy an "invertibility" condition. This second outer bound now reduces to the capacity of this special class of deterministic interference channels. The second outer bound is further tightened for the high SNR deterministic approximation of the Gaussian interference channel with a cognitive relay by exploiting the special structure of the interference. We provide an example that suggests that this third bound is tight in at least some parameter regimes for the high SNR deterministic approximation of the Gaussian channel. Another example shows that the third bound is capacity in the special case where there are no direct links between the non-cognitive transmitters.
1004.4949
Reed Muller Sensing Matrices and the LASSO
cs.IT math.IT
We construct two families of deterministic sensing matrices where the columns are obtained by exponentiating codewords in the quaternary Delsarte-Goethals code $DG(m,r)$. This method of construction results in sensing matrices with low coherence and spectral norm. The first family, which we call Delsarte-Goethals frames, are $2^m$ - dimensional tight frames with redundancy $2^{rm}$. The second family, which we call Delsarte-Goethals sieves, are obtained by subsampling the column vectors in a Delsarte-Goethals frame. Different rows of a Delsarte-Goethals sieve may not be orthogonal, and we present an effective algorithm for identifying all pairs of non-orthogonal rows. The pairs turn out to be duplicate measurements and eliminating them leads to a tight frame. Experimental results suggest that all $DG(m,r)$ sieves with $m\leq 15$ and $r\geq2$ are tight-frames; there are no duplicate rows. For both families of sensing matrices, we measure accuracy of reconstruction (statistical 0-1 loss) and complexity (average reconstruction time) as a function of the sparsity level $k$. Our results show that DG frames and sieves outperform random Gaussian matrices in terms of noiseless and noisy signal recovery using the LASSO.
1004.4965
Many-to-Many Graph Matching: a Continuous Relaxation Approach
stat.ML cs.CV
Graphs provide an efficient tool for object representation in various computer vision applications. Once graph-based representations are constructed, an important question is how to compare graphs. This problem is often formulated as a graph matching problem where one seeks a mapping between vertices of two graphs which optimally aligns their structure. In the classical formulation of graph matching, only one-to-one correspondences between vertices are considered. However, in many applications, graphs cannot be matched perfectly and it is more interesting to consider many-to-many correspondences where clusters of vertices in one graph are matched to clusters of vertices in the other graph. In this paper, we formulate the many-to-many graph matching problem as a discrete optimization problem and propose an approximate algorithm based on a continuous relaxation of the combinatorial problem. We compare our method with other existing methods on several benchmark computer vision datasets.
1004.4968
On the Achievable Rate Regions for a Class of Cognitive Radio Channels: Interference Channel with Degraded Message Sets with Unidirectional Destination Cooperation
cs.IT math.IT
This paper considers the capacity gains due to unidirectional destination cooperation in cognitive radio channels. We propose a novel channel, interference channel with degraded message sets with unidirectional destination cooperation (IC-DMS-UDC), to allow the receiver of cognitive radio (secondary user) to participate in relaying the information for primary system (legitimate user). Our main result is the development of an achievable rate region which combines Gel'fand-Pinkser coding with partial-decode-and-forward strategy employed in the relay channel. A numerical evaluation of the region in the Gaussian case is also provided to demonstrate the improvements.
1004.5026
Compressed Sensing: How sharp is the Restricted Isometry Property
cs.IT math.IT
Compressed Sensing (CS) seeks to recover an unknown vector with $N$ entries by making far fewer than $N$ measurements; it posits that the number of compressed sensing measurements should be comparable to the information content of the vector, not simply $N$. CS combines the important task of compression directly with the measurement task. Since its introduction in 2004 there have been hundreds of manuscripts on CS, a large fraction of which develop algorithms to recover a signal from its compressed measurements. Because of the paradoxical nature of CS -- exact reconstruction from seemingly undersampled measurements -- it is crucial for acceptance of an algorithm that rigorous analyses verify the degree of undersampling the algorithm permits. The Restricted Isometry Property (RIP) has become the dominant tool used for the analysis in such cases. We present here an asymmetric form of RIP which gives tighter bounds than the usual symmetric one. We give the best known bounds on the RIP constants for matrices from the Gaussian ensemble. Our derivations illustrate the way in which the combinatorial nature of CS is controlled. Our quantitative bounds on the RIP allow precise statements as to how aggressively a signal can be undersampled, the essential question for practitioners. We also document the extent to which RIP gives precise information about the true performance limits of CS, by comparing with approaches from high-dimensional geometry.
1004.5049
The Burbea-Rao and Bhattacharyya centroids
cs.IT cs.CG math.IT
We study the centroid with respect to the class of information-theoretic Burbea-Rao divergences that generalize the celebrated Jensen-Shannon divergence by measuring the non-negative Jensen difference induced by a strictly convex and differentiable function. Although those Burbea-Rao divergences are symmetric by construction, they are not metric since they fail to satisfy the triangle inequality. We first explain how a particular symmetrization of Bregman divergences called Jensen-Bregman distances yields exactly those Burbea-Rao divergences. We then proceed by defining skew Burbea-Rao divergences, and show that skew Burbea-Rao divergences amount in limit cases to compute Bregman divergences. We then prove that Burbea-Rao centroids are unique, and can be arbitrarily finely approximated by a generic iterative concave-convex optimization algorithm with guaranteed convergence property. In the second part of the paper, we consider the Bhattacharyya distance that is commonly used to measure overlapping degree of probability distributions. We show that Bhattacharyya distances on members of the same statistical exponential family amount to calculate a Burbea-Rao divergence in disguise. Thus we get an efficient algorithm for computing the Bhattacharyya centroid of a set of parametric distributions belonging to the same exponential families, improving over former specialized methods found in the literature that were limited to univariate or "diagonal" multivariate Gaussians. To illustrate the performance of our Bhattacharyya/Burbea-Rao centroid algorithm, we present experimental performance results for $k$-means and hierarchical clustering methods of Gaussian mixture models.
1004.5051
Tailored RF pulse optimization for magnetization inversion at ultra high field
cs.CE cs.NE physics.med-ph
The radiofrequency (RF) transmit field is severely inhomogeneous at ultrahigh field due to both RF penetration and RF coil design issues. This particularly impairs image quality for sequences that use inversion pulses such as magnetization prepared rapid acquisition gradient echo and limits the use of quantitative arterial spin labeling sequences such as flow-attenuated inversion recovery. Here we have used a search algorithm to produce inversion pulses tailored to take into account the heterogeneity of the RF transmit field at 7 T. This created a slice selective inversion pulse that worked well (good slice profile and uniform inversion) over the range of RF amplitudes typically obtained in the head at 7 T while still maintaining an experimentally achievable pulse length and pulse amplitude in the brain at 7 T. The pulses used were based on the frequency offset correction inversion technique, as well as time dilation of functions, but the RF amplitude, frequency sweep, and gradient functions were all generated using a genetic algorithm with an evaluation function that took into account both the desired inversion profile and the transmit field inhomogeneity.
1004.5070
Multichannel Sampling of Pulse Streams at the Rate of Innovation
cs.IT math.IT
We consider minimal-rate sampling schemes for infinite streams of delayed and weighted versions of a known pulse shape. The minimal sampling rate for these parametric signals is referred to as the rate of innovation and is equal to the number of degrees of freedom per unit time. Although sampling of infinite pulse streams was treated in previous works, either the rate of innovation was not achieved, or the pulse shape was limited to Diracs. In this paper we propose a multichannel architecture for sampling pulse streams with arbitrary shape, operating at the rate of innovation. Our approach is based on modulating the input signal with a set of properly chosen waveforms, followed by a bank of integrators. This architecture is motivated by recent work on sub-Nyquist sampling of multiband signals. We show that the pulse stream can be recovered from the proposed minimal-rate samples using standard tools taken from spectral estimation in a stable way even at high rates of innovation. In addition, we address practical implementation issues, such as reduction of hardware complexity and immunity to failure in the sampling channels. The resulting scheme is flexible and exhibits better noise robustness than previous approaches.
1004.5071
Dimensions of Formality: A Case Study for MKM in Software Engineering
cs.DL cs.AI cs.SE
We study the formalization of a collection of documents created for a Software Engineering project from an MKM perspective. We analyze how document and collection markup formats can cope with an open-ended, multi-dimensional space of primary and secondary classifications and relationships. We show that RDFa-based extensions of MKM formats, employing flexible "metadata" relationships referencing specific vocabularies for distinct dimensions, are well-suited to encode this and to put it into service. This formalized knowledge can be used for enriching interactive document browsing, for enabling multi-dimensional metadata queries over documents and collections, and for exporting Linked Data to the Semantic Web and thus enabling further reuse.
1004.5094
Fastest Distributed Consensus Problem on Branches of an Arbitrary Connected Sensor Network
cs.IT cs.DC cs.DM math.IT
This paper studies the fastest distributed consensus averaging problem on branches of an arbitrary connected sensor network. In the previous works full knowledge about the sensor network's connectivity topology was required for determining the optimal weights and convergence rate of distributed consensus averaging algorithm over the network. Here in this work for the first time, the optimal weights are determined analytically for the edges of certain types of branches, independent of the rest of network. The solution procedure consists of stratification of associated connectivity graph of the branches and Semidefinite Programming (SDP), particularly solving the slackness conditions, where the optimal weights are obtained by inductive comparing of the characteristic polynomials initiated by slackness conditions. Several examples and numerical results are provided to confirm the optimality of the obtained weights.
1004.5108
Analyzing Random Network Coding with Differential Equations and Differential Inclusions
cs.IT cs.NI math.DS math.IT
We develop a framework based on differential equations (DE) and differential inclusions (DI) for analyzing Random Network Coding (RNC), as well as a nonlinear variant referred to as Random Coupon (RC), in a wireless network. The DEDI framework serves as a powerful numerical and analytical tool to study RNC. We demonstrate its versatility by proving theoretical results on multicast information flows in a wireless network using RNC or RC. We also demonstrate the accuracy and flexibility of the performance analysis enabled by this framework via illustrative examples of networks with multiple multicast sessions, user cooperation and arbitrary topologies.
1004.5132
The Two-User Deterministic Interference Channel with Rate-Limited Feedback
cs.IT math.IT
In this paper we study the effect of rate-limited feedback on the sum-rate capacity of the deterministic interference channel. We characterize the sum-rate capacity of this channel in the symmetric case and show that having feedback links can increase the sum-rate capacity by at most the rate of the available feedback. Our proof includes a novel upper-bound on the sum-rate capacity and a set of new achievability strategies.
1004.5157
Deriving Good LDPC Convolutional Codes from LDPC Block Codes
cs.IT math.IT
Low-density parity-check (LDPC) convolutional codes are capable of achieving excellent performance with low encoding and decoding complexity. In this paper we discuss several graph-cover-based methods for deriving families of time-invariant and time-varying LDPC convolutional codes from LDPC block codes and show how earlier proposed LDPC convolutional code constructions can be presented within this framework. Some of the constructed convolutional codes significantly outperform the underlying LDPC block codes. We investigate some possible reasons for this "convolutional gain," and we also discuss the --- mostly moderate --- decoder cost increase that is incurred by going from LDPC block to LDPC convolutional codes.
1004.5168
Efficient and Effective Spam Filtering and Re-ranking for Large Web Datasets
cs.IR
The TREC 2009 web ad hoc and relevance feedback tasks used a new document collection, the ClueWeb09 dataset, which was crawled from the general Web in early 2009. This dataset contains 1 billion web pages, a substantial fraction of which are spam --- pages designed to deceive search engines so as to deliver an unwanted payload. We examine the effect of spam on the results of the TREC 2009 web ad hoc and relevance feedback tasks, which used the ClueWeb09 dataset. We show that a simple content-based classifier with minimal training is efficient enough to rank the "spamminess" of every page in the dataset using a standard personal computer in 48 hours, and effective enough to yield significant and substantive improvements in the fixed-cutoff precision (estP10) as well as rank measures (estR-Precision, StatMAP, MAP) of nearly all submitted runs. Moreover, using a set of "honeypot" queries the labeling of training data may be reduced to an entirely automatic process. The results of classical information retrieval methods are particularly enhanced by filtering --- from among the worst to among the best.
1004.5181
Analysis of Feedback Overhead for MIMO Beamforming over Time-Varying Channels
cs.IT math.IT
In this paper, the required amount of feedback overhead for multiple-input multiple-output (MIMO) beamforming over time-varying channels is presented in terms of the entropy of the feedback messages. In the case that each transmit antenna has its own power amplifier which has individual power limit, it has been known that only phase steering information is necessary to form the optimal transmit beamforming vector. Since temporal correlation exists for wireless fading channels, one can utilize the previous reported feedback messages as prior information to efficiently encode the current feedback message. Thus, phase tracking information, difference between two phase steering information in adjacent feedback slots, is sufficient as a feedback message. We show that while the entropy of the phase steering information is a constant, the entropy of the phase tracking information is a function of the temporal correlation parameter. For the phase tracking information, upperbounds on the entropy are presented in the Gaussian entropy and the von-Mises entropy by using the theory on the maximum entropy distributions. Derived results can quantify the amount of reduction in feedback overhead of the phase tracking information over the phase steering information. For application perspective, the signal-to-noise ratio (SNR) gain of phase tracking beamforming over phase steering beamforming is evaluated by using Monte-Carlo simulation. Also we show that the derived entropies can determine the appropriate duration of the feedback reports with respect to the degree of the channel variation rates.
1004.5189
Rate-distortion function via minimum mean square error estimation
cs.IT math.IT
We derive a simple general parametric representation of the rate-distortion function of a memoryless source, where both the rate and the distortion are given by integrals whose integrands include the minimum mean square error (MMSE) of the distortion $\Delta=d(X,Y)$ based on the source symbol $X$, with respect to a certain joint distribution of these two random variables. At first glance, these relations may seem somewhat similar to the I-MMSE relations due to Guo, Shamai and Verd\'u, but they are, in fact, quite different. The new relations among rate, distortion, and MMSE are discussed from several aspects, and more importantly, it is demonstrated that they can sometimes be rather useful for obtaining non-trivial upper and lower bounds on the rate-distortion function, as well as for determining the exact asymptotic behavior for very low and for very large distortion. Analogous MMSE relations hold for channel capacity as well.
1004.5194
Clustering processes
cs.LG cs.IT math.IT stat.ML
The problem of clustering is considered, for the case when each data point is a sample generated by a stationary ergodic process. We propose a very natural asymptotic notion of consistency, and show that simple consistent algorithms exist, under most general non-parametric assumptions. The notion of consistency is as follows: two samples should be put into the same cluster if and only if they were generated by the same distribution. With this notion of consistency, clustering generalizes such classical statistical problems as homogeneity testing and process classification. We show that, for the case of a known number of clusters, consistency can be achieved under the only assumption that the joint distribution of the data is stationary ergodic (no parametric or Markovian assumptions, no assumptions of independence, neither between nor within the samples). If the number of clusters is unknown, consistency can be achieved under appropriate assumptions on the mixing rates of the processes. (again, no parametric or independence assumptions). In both cases we give examples of simple (at most quadratic in each argument) algorithms which are consistent.
1004.5195
On Perfect Codes in the Johnson Graph
cs.IT math.IT
In this paper we consider the existence of nontrivial perfect codes in the Johnson graph J(n,w). We present combinatorial and number theory techniques to provide necessary conditions for existence of such codes and reduce the range of parameters in which 1-perfect and 2-perfect codes may exist.
1004.5214
Split-Extended LDPC codes for coded cooperation
cs.IT math.IT
We propose a new code design that aims to distribute an LDPC code over a relay channel. It is based on a split-and-extend approach, which allows the relay to split the set of bits connected to some parity-check of the LDPC code into two or several subsets. Subsequently, the sums of bits within each subset are used in a repeat-accumulate manner in order to generate extra bits sent from the relay toward the destination. We show that the proposed design yields LDPC codes with enhanced correction capacity and can be advantageously applied to existing codes, which allows for addressing cooperation issues for evolving standards. Finally, we derive density evolution equations for the proposed design, and we show that Split-Extended LDPC codes can approach very closely the capacity of the Gaussian relay channel.
1004.5215
System Dynamics Modelling of the Processes Involving the Maintenance of the Naive T Cell Repertoire
cs.AI q-bio.CB
The study of immune system aging, i.e. immunosenescence, is a relatively new research topic. It deals with understanding the processes of immunodegradation that indicate signs of functionality loss possibly leading to death. Even though it is not possible to prevent immunosenescence, there is great benefit in comprehending its causes, which may help to reverse some of the damage done and thus improve life expectancy. One of the main factors influencing the process of immunosenescence is the number and phenotypical variety of naive T cells in an individual. This work presents a review of immunosenescence, proposes system dynamics modelling of the processes involving the maintenance of the naive T cell repertoire and presents some preliminary results.
1004.5216
Optimized puncturing distributions for irregular non-binary LDPC codes
cs.IT math.IT
In this paper we design non-uniform bit-wise puncturing distributions for irregular non-binary LDPC (NB-LDPC) codes. The puncturing distributions are optimized by minimizing the decoding threshold of the punctured LDPC code, the threshold being computed with a Monte-Carlo implementation of Density Evolution. First, we show that Density Evolution computed with Monte-Carlo simulations provides accurate (very close) and precise (small variance) estimates of NB-LDPC code ensemble thresholds. Based on the proposed method, we analyze several puncturing distributions for regular and semi-regular codes, obtained either by clustering punctured bits, or spreading them over the symbol-nodes of the Tanner graph. Finally, optimized puncturing distributions for non-binary LDPC codes with small maximum degree are presented, which exhibit a gap between 0.2 and 0.5 dB to the channel capacity, for punctured rates varying from 0.5 to 0.9.
1004.5217
Analysis of Quasi-Cyclic LDPC codes under ML decoding over the erasure channel
cs.IT math.IT
In this paper, we show that Quasi-Cyclic LDPC codes can efficiently accommodate the hybrid iterative/ML decoding over the binary erasure channel. We demonstrate that the quasi-cyclic structure of the parity-check matrix can be advantageously used in order to significantly reduce the complexity of the ML decoding. This is achieved by a simple row/column permutation that transforms a QC matrix into a pseudo-band form. Based on this approach, we propose a class of QC-LDPC codes with almost ideal error correction performance under the ML decoding, while the required number of row/symbol operations scales as $k\sqrt{k}$, where $k$ is the number of source symbols.
1004.5222
The Application of a Dendritic Cell Algorithm to a Robotic Classifier
cs.AI cs.NE cs.RO
The dendritic cell algorithm is an immune-inspired technique for processing time-dependant data. Here we propose it as a possible solution for a robotic classification problem. The dendritic cell algorithm is implemented on a real robot and an investigation is performed into the effects of varying the migration threshold median for the cell population. The algorithm performs well on a classification task with very little tuning. Ways of extending the implementation to allow it to be used as a classifier within the field of robotic security are suggested.
1004.5229
Optimism in Reinforcement Learning and Kullback-Leibler Divergence
cs.LG math.ST stat.ML stat.TH
We consider model-based reinforcement learning in finite Markov De- cision Processes (MDPs), focussing on so-called optimistic strategies. In MDPs, optimism can be implemented by carrying out extended value it- erations under a constraint of consistency with the estimated model tran- sition probabilities. The UCRL2 algorithm by Auer, Jaksch and Ortner (2009), which follows this strategy, has recently been shown to guarantee near-optimal regret bounds. In this paper, we strongly argue in favor of using the Kullback-Leibler (KL) divergence for this purpose. By studying the linear maximization problem under KL constraints, we provide an ef- ficient algorithm, termed KL-UCRL, for solving KL-optimistic extended value iteration. Using recent deviation bounds on the KL divergence, we prove that KL-UCRL provides the same guarantees as UCRL2 in terms of regret. However, numerical experiments on classical benchmarks show a significantly improved behavior, particularly when the MDP has reduced connectivity. To support this observation, we provide elements of com- parison between the two algorithms based on geometric considerations.
1004.5262
On Application of the Local Search and the Genetic Algorithms Techniques to Some Combinatorial Optimization Problems
cs.NE math.OC
In this paper the approach to solving several combinatorial optimization problems using the local search and the genetic algorithm techniques is proposed. Initially this approach was developed in purpose to overcome some difficulties inhibiting the application of above mentioned techniques to the problems of the Questionnaire Theory. But when the algorithms were developed it became clear that them could be successfully applied also to the Minimum Set Cover, the 0-1-Knapsack and probably to other combinatorial optimization problems.
1004.5274
Robustness maximization of parallel multichannel systems
cs.IT math.IT
Bit error rate (BER) minimization and SNR-gap maximization, two robustness optimization problems, are solved, under average power and bit-rate constraints, according to the waterfilling policy. Under peak-power constraint the solutions differ and this paper gives bit-loading solutions of both robustness optimization problems over independent parallel channels. The study is based on analytical approach with generalized Lagrangian relaxation tool and on greedy-type algorithm approach. Tight BER expressions are used for square and rectangular quadrature amplitude modulations. Integer bit solution of analytical continuous bit-rates is performed with a new generalized secant method. The asymptotic convergence of both robustness optimizations is proved for both analytical and algorithmic approaches. We also prove that, in conventional margin maximization problem, the equivalence between SNR-gap maximization and power minimization does not hold with peak-power limitation. Based on a defined dissimilarity measure, bit-loading solutions are compared over power line communication channel for multicarrier systems. Simulation results confirm the asymptotic convergence of both allocation policies. In non asymptotic regime the allocation policies can be interchanged depending on the robustness measure and the operating point of the communication system. The low computational effort of the suboptimal solution based on analytical approach leads to a good trade-off between performance and complexity.
1004.5305
Compressed Sensing with off-axis frequency-shifting holography
physics.optics cs.CV physics.med-ph
This work reveals an experimental microscopy acquisition scheme successfully combining Compressed Sensing (CS) and digital holography in off-axis and frequency-shifting conditions. CS is a recent data acquisition theory involving signal reconstruction from randomly undersampled measurements, exploiting the fact that most images present some compact structure and redundancy. We propose a genuine CS-based imaging scheme for sparse gradient images, acquiring a diffraction map of the optical field with holographic microscopy and recovering the signal from as little as 7% of random measurements. We report experimental results demonstrating how CS can lead to an elegant and effective way to reconstruct images, opening the door for new microscopy applications.
1004.5326
Designing neural networks that process mean values of random variables
cond-mat.dis-nn cs.AI cs.LG
We introduce a class of neural networks derived from probabilistic models in the form of Bayesian networks. By imposing additional assumptions about the nature of the probabilistic models represented in the networks, we derive neural networks with standard dynamics that require no training to determine the synaptic weights, that perform accurate calculation of the mean values of the random variables, that can pool multiple sources of evidence, and that deal cleanly and consistently with inconsistent or contradictory evidence. The presented neural networks capture many properties of Bayesian networks, providing distributed versions of probabilistic models.
1004.5339
Query strategy for sequential ontology debugging
cs.LO cs.AI
Debugging of ontologies is an important prerequisite for their wide-spread application, especially in areas that rely upon everyday users to create and maintain knowledge bases, as in the case of the Semantic Web. Recent approaches use diagnosis methods to identify causes of inconsistent or incoherent ontologies. However, in most debugging scenarios these methods return many alternative diagnoses, thus placing the burden of fault localization on the user. This paper demonstrates how the target diagnosis can be identified by performing a sequence of observations, that is, by querying an oracle about entailments of the target ontology. We exploit a-priori probabilities of typical user errors to formulate information-theoretic concepts for query selection. Our evaluation showed that the proposed method significantly reduces the number of required queries compared to myopic strategies. We experimented with different probability distributions of user errors and different qualities of the a-priori probabilities. Our measurements showed the advantageousness of information-theoretic approach to query selection even in cases where only a rough estimate of the priors is available.
1004.5351
Isometric Embeddings in Imaging and Vision: Facts and Fiction
cs.CV math.CV math.DG
We explore the practicability of Nash's Embedding Theorem in vision and imaging sciences. In particular, we investigate the relevance of a result of Burago and Zalgaller regarding the existence of isometric embeddings of polyhedral surfaces in $\mathbb{R}^3$ and we show that their proof does not extended directly to higher dimensions.
1004.5367
Multiplicatively Repeated Non-Binary LDPC Codes
cs.IT math.IT
We propose non-binary LDPC codes concatenated with multiplicative repetition codes. By multiplicatively repeating the (2,3)-regular non-binary LDPC mother code of rate 1/3, we construct rate-compatible codes of lower rates 1/6, 1/9, 1/12,... Surprisingly, such simple low-rate non-binary LDPC codes outperform the best low-rate binary LDPC codes so far. Moreover, we propose the decoding algorithm for the proposed codes, which can be decoded with almost the same computational complexity as that of the mother code.
1004.5370
Self-Taught Hashing for Fast Similarity Search
cs.IR
The ability of fast similarity search at large scale is of great importance to many Information Retrieval (IR) applications. A promising way to accelerate similarity search is semantic hashing which designs compact binary codes for a large number of documents so that semantically similar documents are mapped to similar codes (within a short Hamming distance). Although some recently proposed techniques are able to generate high-quality codes for documents known in advance, obtaining the codes for previously unseen documents remains to be a very challenging problem. In this paper, we emphasise this issue and propose a novel Self-Taught Hashing (STH) approach to semantic hashing: we first find the optimal $l$-bit binary codes for all documents in the given corpus via unsupervised learning, and then train $l$ classifiers via supervised learning to predict the $l$-bit code for any query document unseen before. Our experiments on three real-world text datasets show that the proposed approach using binarised Laplacian Eigenmap (LapEig) and linear Support Vector Machine (SVM) outperforms state-of-the-art techniques significantly.
1004.5421
Interference Mitigation through Limited Transmitter Cooperation
cs.IT math.IT
Interference limits performance in wireless networks, and cooperation among receivers or transmitters can help mitigate interference by forming distributed MIMO systems. Earlier work shows how limited receiver cooperation helps mitigate interference. The scenario with transmitter cooperation, however, is more difficult to tackle. In this paper we study the two-user Gaussian interference channel with conferencing transmitters to make progress towards this direction. We characterize the capacity region to within 6.5 bits/s/Hz, regardless of channel parameters. Based on the constant-to-optimality result, we show that there is an interesting reciprocity between the scenario with conferencing transmitters and the scenario with conferencing receivers, and their capacity regions are within a constant gap to each other. Hence in the interference-limited regime, the behavior of the benefit brought by transmitter cooperation is the same as that by receiver cooperation.
1004.5424
Graphic Symbol Recognition using Graph Based Signature and Bayesian Network Classifier
cs.CV cs.GR
We present a new approach for recognition of complex graphic symbols in technical documents. Graphic symbol recognition is a well known challenge in the field of document image analysis and is at heart of most graphic recognition systems. Our method uses structural approach for symbol representation and statistical classifier for symbol recognition. In our system we represent symbols by their graph based signatures: a graphic symbol is vectorized and is converted to an attributed relational graph, which is used for computing a feature vector for the symbol. This signature corresponds to geometry and topology of the symbol. We learn a Bayesian network to encode joint probability distribution of symbol signatures and use it in a supervised learning scenario for graphic symbol recognition. We have evaluated our method on synthetically deformed and degraded images of pre-segmented 2D architectural and electronic symbols from GREC databases and have obtained encouraging recognition rates.
1004.5427
Employing fuzzy intervals and loop-based methodology for designing structural signature: an application to symbol recognition
cs.CV cs.GR
Motivation of our work is to present a new methodology for symbol recognition. We support structural methods for representing visual associations in graphic documents. The proposed method employs a structural approach for symbol representation and a statistical classifier for recognition. We vectorize a graphic symbol, encode its topological and geometrical information by an ARG and compute a signature from this structural graph. To address the sensitivity of structural representations to deformations and degradations, we use data adapted fuzzy intervals while computing structural signature. The joint probability distribution of signatures is encoded by a Bayesian network. This network in fact serves as a mechanism for pruning irrelevant features and choosing a subset of interesting features from structural signatures, for underlying symbol set. Finally we deploy the Bayesian network in supervised learning scenario for recognizing query symbols. We have evaluated the robustness of our method against noise, on synthetically deformed and degraded images of pre-segmented 2D architectural and electronic symbols from GREC databases and have obtained encouraging recognition rates. A second set of experimentation was carried out for evaluating the performance of our method against context noise i.e. symbols cropped from complete documents. The results support the use of our signature by a symbol spotting system.
1004.5429
On Distance Properties of Quasi-Cyclic Protograph-Based LDPC Codes
cs.IT math.IT
Recent work has shown that properly designed protograph-based LDPC codes may have minimum distance linearly increasing with block length. This notion rests on ensemble arguments over all possible expansions of the base protograph. When implementation complexity is considered, the expansion is typically chosen to be quite orderly. For example, protograph expansion by cyclically shifting connections creates a quasi-cyclic (QC) code. Other recent work has provided upper bounds on the minimum distance of QC codes. In this paper, these bounds are expanded upon to cover puncturing and tightened in several specific cases. We then evaluate our upper bounds for the most prominent protograph code thus far, one proposed for deep-space usage in the CCSDS experimental standard, the code known as AR4JA.
1004.5442
Multiple-Relaxation-Time Lattice Boltzmann Approach to Compressible Flows with Flexible Specific-Heat Ratio and Prandtl Number
cond-mat.soft cs.CE nlin.CG physics.comp-ph physics.flu-dyn stat.CO
A new multiple-relaxation-time lattice Boltzmann scheme for compressible flows with arbitrary specific heat ratio and Prandtl number is presented. In the new scheme, which is based on a two-dimensional 16-discrete-velocity model, the moment space and the corresponding transformation matrix are constructed according to the seven-moment relations associated with the local equilibrium distribution function. In the continuum limit, the model recovers the compressible Navier-Stokes equations with flexible specific-heat ratio and Prandtl number. Numerical experiments show that compressible flows with strong shocks can be simulated by the present model up to Mach numbers $Ma \sim 5$.
1004.5479
On Minimax Robust Detection of Stationary Gaussian Signals in White Gaussian Noise
cs.IT math.IT math.ST stat.TH
The problem of detecting a wide-sense stationary Gaussian signal process embedded in white Gaussian noise, where the power spectral density of the signal process exhibits uncertainty, is investigated. The performance of minimax robust detection is characterized by the exponential decay rate of the miss probability under a Neyman-Pearson criterion with a fixed false alarm probability, as the length of the observation interval grows without bound. A dominance condition is identified for the uncertainty set of spectral density functions, and it is established that, under the dominance condition, the resulting minimax problem possesses a saddle point, which is achievable by the likelihood ratio tests matched to a so-called dominated power spectral density in the uncertainty set. No convexity condition on the uncertainty set is required to establish this result.
1004.5500
Simple Type Theory as Framework for Combining Logics
cs.LO cs.AI
Simple type theory is suited as framework for combining classical and non-classical logics. This claim is based on the observation that various prominent logics, including (quantified) multimodal logics and intuitionistic logics, can be elegantly embedded in simple type theory. Furthermore, simple type theory is sufficiently expressive to model combinations of embedded logics and it has a well understood semantics. Off-the-shelf reasoning systems for simple type theory exist that can be uniformly employed for reasoning within and about combinations of logics.
1004.5529
High-Rate Vector Quantization for the Neyman-Pearson Detection of Correlated Processes
cs.IT math.IT math.PR math.ST stat.TH
This paper investigates the effect of quantization on the performance of the Neyman-Pearson test. It is assumed that a sensing unit observes samples of a correlated stationary ergodic multivariate process. Each sample is passed through an N-point quantizer and transmitted to a decision device which performs a binary hypothesis test. For any false alarm level, it is shown that the miss probability of the Neyman-Pearson test converges to zero exponentially as the number of samples tends to infinity, assuming that the observed process satisfies certain mixing conditions. The main contribution of this paper is to provide a compact closed-form expression of the error exponent in the high-rate regime i.e., when the number N of quantization levels tends to infinity, generalizing previous results of Gupta and Hero to the case of non-independent observations. If d represents the dimension of one sample, it is proved that the error exponent converges at rate N^{2/d} to the one obtained in the absence of quantization. As an application, relevant high-rate quantization strategies which lead to a large error exponent are determined. Numerical results indicate that the proposed quantization rule can yield better performance than existing ones in terms of detection error.
1004.5538
Bayesian estimation of regularization and PSF parameters for Wiener-Hunt deconvolution
stat.CO cs.CV physics.data-an stat.ME
This paper tackles the problem of image deconvolution with joint estimation of PSF parameters and hyperparameters. Within a Bayesian framework, the solution is inferred via a global a posteriori law for unknown parameters and object. The estimate is chosen as the posterior mean, numerically calculated by means of a Monte-Carlo Markov chain algorithm. The estimates are efficiently computed in the Fourier domain and the effectiveness of the method is shown on simulated examples. Results show precise estimates for PSF parameters and hyperparameters as well as precise image estimates including restoration of high-frequencies and spatial details, within a global and coherent approach.
1004.5540
Strong Secrecy for Erasure Wiretap Channels
cs.IT math.IT
We show that duals of certain low-density parity-check (LDPC) codes, when used in a standard coset coding scheme, provide strong secrecy over the binary erasure wiretap channel (BEWC). This result hinges on a stopping set analysis of ensembles of LDPC codes with block length $n$ and girth $\geq 2k$, for some $k \geq 2$. We show that if the minimum left degree of the ensemble is $l_\mathrm{min}$, the expected probability of block error is $\calO(\frac{1}{n^{\lceil l_\mathrm{min} k /2 \rceil - k}})$ when the erasure probability $\epsilon < \epsilon_\mathrm{ef}$, where $\epsilon_\mathrm{ef}$ depends on the degree distribution of the ensemble. As long as $l_\mathrm{min} > 2$ and $k > 2$, the dual of this LDPC code provides strong secrecy over a BEWC of erasure probability greater than $1 - \epsilon_\mathrm{ef}$.
1004.5551
Entanglement Transmission over Arbitrarily Varying Quantum Channels
quant-ph cs.IT math.IT
We derive a regularized formula for the common randomness assisted entanglement transmission capacity of finite arbitrarily varying quantum channels (AVQC's). For finite AVQC's with positive capacity for classical message transmission we show, by derandomization through classical forward communication, that the random capacity for entanglement transmission equals the deterministic capacity for entanglement transmission. This is a quantum version of the famous Ahlswede dichotomy. In the infinite case, we derive a similar result for certain classes of AVQC's. At last, we give two possible definitions of symmetrizability of an AVQC.
1004.5570
Optimal computation of symmetric Boolean functions in Tree networks
cs.IT cs.NI math.IT
In this paper, we address the scenario where nodes with sensor data are connected in a tree network, and every node wants to compute a given symmetric Boolean function of the sensor data. We first consider the problem of computing a function of two nodes with integer measurements. We allow for block computation to enhance data fusion efficiency, and determine the minimum worst-case total number of bits to be exchanged to perform the desired computation. We establish lower bounds using fooling sets, and provide a novel scheme which attains the lower bounds, using information theoretic tools. For a class of functions called sum-threshold functions, this scheme is shown to be optimal. We then turn to tree networks and derive a lower bound for the number of bits exchanged on each link by viewing it as a two node problem. We show that the protocol of recursive innetwork aggregation achieves this lower bound in the case of sumthreshold functions. Thus we have provided a communication and in-network computation strategy that is optimal for each link. All the results can be extended to the case of non-binary alphabets. In the case of general graphs, we present a cut-set lower bound, and an achievable scheme based on aggregation along trees. For complete graphs, the complexity of this scheme is no more than twice that of the optimal scheme.
1004.5571
Optimal ordering of transmissions for computing Boolean threhold functions
cs.IT cs.NI math.IT
We address a sequential decision problem that arises in the computation of symmetric Boolean functions of distributed data. We consider a collocated network, where each node's transmissions can be heard by every other node. Each node has a Boolean measurement and we wish to compute a given Boolean function of these measurements. We suppose that the measurements are independent and Bernoulli distributed. Thus, the problem of optimal computation becomes the problem of optimally ordering node's transmissions so as to minimize the total expected number of bits. We solve the ordering problem for the class of Boolean threshold functions. The optimal ordering is dynamic, i.e., it could potentially depend on the values of previously transmitted bits. Further, it depends only on the ordering of the marginal probabilites, but not on their exact values. This provides an elegant structure for the optimal strategy. For the case where each node has a block of measurements, the problem is significantly harder, and we conjecture the optimal strategy.
1004.5588
On Achieving Local View Capacity Via Maximal Independent Graph Scheduling
cs.IT math.IT
"If we know more, we can achieve more." This adage also applies to communication networks, where more information about the network state translates into higher sumrates. In this paper, we formalize this increase of sum-rate with increased knowledge of the network state. The knowledge of network state is measured in terms of the number of hops, h, of information available to each transmitter and is labeled as h-local view. To understand how much capacity is lost due to limited information, we propose to use the metric of normalized sum-capacity, which is the h-local view sum-capacity divided by global-view sum capacity. For the cases of one and two-local view, we characterize the normalized sum-capacity for many classes of deterministic and Gaussian interference networks. In many cases, a scheduling scheme called maximal independent graph scheduling is shown to achieve normalized sum-capacity. We also show that its generalization for 1-local view, labeled coded set scheduling, achieves normalized sum-capacity in some cases where its uncoded counterpart fails to do so.
1004.5601
Near MDS poset codes and distributions
cs.IT math.IT
We study $q$-ary codes with distance defined by a partial order of the coordinates of the codewords. Maximum Distance Separable (MDS) codes in the poset metric have been studied in a number of earlier works. We consider codes that are close to MDS codes by the value of their minimum distance. For such codes, we determine their weight distribution, and in the particular case of the "ordered metric" characterize distributions of points in the unit cube defined by the codes. We also give some constructions of codes in the ordered Hamming space.
1005.0027
Learning from Multiple Outlooks
cs.LG
We propose a novel problem formulation of learning a single task when the data are provided in different feature spaces. Each such space is called an outlook, and is assumed to contain both labeled and unlabeled data. The objective is to take advantage of the data from all the outlooks to better classify each of the outlooks. We devise an algorithm that computes optimal affine mappings from different outlooks to a target outlook by matching moments of the empirical distributions. We further derive a probabilistic interpretation of the resulting algorithm and a sample complexity bound indicating how many samples are needed to adequately find the mapping. We report the results of extensive experiments on activity recognition tasks that show the value of the proposed approach in boosting performance.
1005.0047
A Geometric View of Conjugate Priors
cs.LG
In Bayesian machine learning, conjugate priors are popular, mostly due to mathematical convenience. In this paper, we show that there are deeper reasons for choosing a conjugate prior. Specifically, we formulate the conjugate prior in the form of Bregman divergence and show that it is the inherent geometry of conjugate priors that makes them appropriate and intuitive. This geometric interpretation allows one to view the hyperparameters of conjugate priors as the {\it effective} sample points, thus providing additional intuition. We use this geometric understanding of conjugate priors to derive the hyperparameters and expression of the prior used to couple the generative and discriminative components of a hybrid model for semi-supervised learning.
1005.0052
On the Joint Decoding of LDPC Codes and Finite-State Channels via Linear Programming
cs.IT math.IT
In this paper, the linear programming (LP) decoder for binary linear codes, introduced by Feldman, et al. is extended to joint-decoding of binary-input finite-state channels. In particular, we provide a rigorous definition of LP joint-decoding pseudo-codewords (JD-PCWs) that enables evaluation of the pairwise error probability between codewords and JD-PCWs. This leads naturally to a provable upper bound on decoder failure probability. If the channel is a finite-state intersymbol interference channel, then the LP joint decoder also has the maximum-likelihood (ML) certificate property and all integer valued solutions are codewords. In this case, the performance loss relative to ML decoding can be explained completely by fractional valued JD-PCWs.
1005.0063
Large Margin Multiclass Gaussian Classification with Differential Privacy
stat.ML cs.CR cs.LG
As increasing amounts of sensitive personal information is aggregated into data repositories, it has become important to develop mechanisms for processing the data without revealing information about individual data instances. The differential privacy model provides a framework for the development and theoretical analysis of such mechanisms. In this paper, we propose an algorithm for learning a discriminatively trained multi-class Gaussian classifier that satisfies differential privacy using a large margin loss function with a perturbed regularization term. We present a theoretical upper bound on the excess risk of the classifier introduced by the perturbation.
1005.0069
Perturbation Resilience and Superiorization of Iterative Algorithms
math.OC cs.CV physics.med-ph
Iterative algorithms aimed at solving some problems are discussed. For certain problems, such as finding a common point in the intersection of a finite number of convex sets, there often exist iterative algorithms that impose very little demand on computer resources. For other problems, such as finding that point in the intersection at which the value of a given function is optimal, algorithms tend to need more computer memory and longer execution time. A methodology is presented whose aim is to produce automatically for an iterative algorithm of the first kind a "superiorized version" of it that retains its computational efficiency but nevertheless goes a long way towards solving an optimization problem. This is possible to do if the original algorithm is "perturbation resilient," which is shown to be the case for various projection algorithms for solving the consistent convex feasibility problem. The superiorized versions of such algorithms use perturbations that drive the process in the direction of the optimizer of the given function. After presenting these intuitive ideas in a precise mathematical form, they are illustrated in image reconstruction from projections for two different projection algorithms superiorized for the function whose value is the total variation of the image.
1005.0072
HyberLoc: Providing Physical Layer Location Privacy in Hybrid Sensor Networks
cs.IT cs.CR math.IT
In many hybrid wireless sensor networks' applications, sensor nodes are deployed in hostile environments where trusted and un-trusted nodes co-exist. In anchor-based hybrid networks, it becomes important to allow trusted nodes to gain full access to the location information transmitted in beacon frames while, at the same time, prevent un-trusted nodes from using this information. The main challenge is that un-trusted nodes can measure the physical signal transmitted from anchor nodes, even if these nodes encrypt their transmission. Using the measured signal strength, un-trusted nodes can still tri-laterate the location of anchor nodes. In this paper, we propose HyberLoc, an algorithm that provides anchor physical layer location privacy in anchor-based hybrid sensor networks. The idea is for anchor nodes to dynamically change their transmission power following a certain probability distribution, degrading the localization accuracy at un-trusted nodes while maintaining high localization accuracy at trusted nodes. Given an average power constraint, our analysis shows that the discretized exponential distribution is the distribution that maximizes location uncertainty at the untrusted nodes. Detailed evaluation through analysis, simulation, and implementation shows that HyberLoc gives trusted nodes up to 3.5 times better localization accuracy as compared to untrusted nodes.
1005.0075
Distributive Stochastic Learning for Delay-Optimal OFDMA Power and Subband Allocation
cs.LG
In this paper, we consider the distributive queue-aware power and subband allocation design for a delay-optimal OFDMA uplink system with one base station, $K$ users and $N_F$ independent subbands. Each mobile has an uplink queue with heterogeneous packet arrivals and delay requirements. We model the problem as an infinite horizon average reward Markov Decision Problem (MDP) where the control actions are functions of the instantaneous Channel State Information (CSI) as well as the joint Queue State Information (QSI). To address the distributive requirement and the issue of exponential memory requirement and computational complexity, we approximate the subband allocation Q-factor by the sum of the per-user subband allocation Q-factor and derive a distributive online stochastic learning algorithm to estimate the per-user Q-factor and the Lagrange multipliers (LM) simultaneously and determine the control actions using an auction mechanism. We show that under the proposed auction mechanism, the distributive online learning converges almost surely (with probability 1). For illustration, we apply the proposed distributive stochastic learning framework to an application example with exponential packet size distribution. We show that the delay-optimal power control has the {\em multi-level water-filling} structure where the CSI determines the instantaneous power allocation and the QSI determines the water-level. The proposed algorithm has linear signaling overhead and computational complexity $\mathcal O(KN)$, which is desirable from an implementation perspective.
1005.0080
Electronic Geometry Textbook: A Geometric Textbook Knowledge Management System
cs.AI cs.MS
Electronic Geometry Textbook is a knowledge management system that manages geometric textbook knowledge to enable users to construct and share dynamic geometry textbooks interactively and efficiently. Based on a knowledge base organizing and storing the knowledge represented in specific languages, the system implements interfaces for maintaining the data representing that knowledge as well as relations among those data, for automatically generating readable documents for viewing or printing, and for automatically discovering the relations among knowledge data. An interface has been developed for users to create geometry textbooks with automatic checking, in real time, of the consistency of the structure of each resulting textbook. By integrating an external geometric theorem prover and an external dynamic geometry software package, the system offers the facilities for automatically proving theorems and generating dynamic figures in the created textbooks. This paper provides a comprehensive account of the current version of Electronic Geometry Textbook.
1005.0089
The Exact Closest String Problem as a Constraint Satisfaction Problem
cs.AI
We report (to our knowledge) the first evaluation of Constraint Satisfaction as a computational framework for solving closest string problems. We show that careful consideration of symbol occurrences can provide search heuristics that provide several orders of magnitude speedup at and above the optimal distance. We also report (to our knowledge) the first analysis and evaluation -- using any technique -- of the computational difficulties involved in the identification of all closest strings for a given input set. We describe algorithms for web-scale distributed solution of closest string problems, both purely based on AI backtrack search and also hybrid numeric-AI methods.
1005.0104
Joint Structured Models for Extraction from Overlapping Sources
cs.AI
We consider the problem of jointly training structured models for extraction from sources whose instances enjoy partial overlap. This has important applications like user-driven ad-hoc information extraction on the web. Such applications present new challenges in terms of the number of sources and their arbitrary pattern of overlap not seen by earlier collective training schemes applied on two sources. We present an agreement-based learning framework and alternatives within it to trade-off tractability, robustness to noise, and extent of agreement. We provide a principled scheme to discover low-noise agreement sets in unlabeled data across the sources. Through extensive experiments over 58 real datasets, we establish that our method of additively rewarding agreement over maximal segments of text provides the best trade-offs, and also scores over alternatives such as collective inference, staged training, and multi-view learning.
1005.0117
On the Separation of Lossy Source-Network Coding and Channel Coding in Wireline Networks
cs.IT math.IT
This paper proves the separation between source-network coding and channel coding in networks of noisy, discrete, memoryless channels. We show that the set of achievable distortion matrices in delivering a family of dependent sources across such a network equals the set of achievable distortion matrices for delivering the same sources across a distinct network which is built by replacing each channel by a noiseless, point-to-point bit-pipe of the corresponding capacity. Thus a code that applies source-network coding across links that are made almost lossless through the application of independent channel coding across each link asymptotically achieves the optimal performance across the network as a whole.
1005.0125
Adaptive Bases for Reinforcement Learning
cs.LG cs.AI
We consider the problem of reinforcement learning using function approximation, where the approximating basis can change dynamically while interacting with the environment. A motivation for such an approach is maximizing the value function fitness to the problem faced. Three errors are considered: approximation square error, Bellman residual, and projected Bellman residual. Algorithms under the actor-critic framework are presented, and shown to converge. The advantage of such an adaptive basis is demonstrated in simulations.
1005.0167
A digital interface for Gaussian relay and interference networks: Lifting codes from the discrete superposition model
cs.IT math.IT
For every Gaussian network, there exists a corresponding deterministic network called the discrete superposition network. We show that this discrete superposition network provides a near-optimal digital interface for operating a class consisting of many Gaussian networks in the sense that any code for the discrete superposition network can be naturally lifted to a corresponding code for the Gaussian network, while achieving a rate that is no more than a constant number of bits lesser than the rate it achieves for the discrete superposition network. This constant depends only on the number of nodes in the network and not on the channel gains or SNR. Moreover the capacities of the two networks are within a constant of each other, again independent of channel gains and SNR. We show that the class of Gaussian networks for which this interface property holds includes relay networks with a single source-destination pair, interference networks, multicast networks, and the counterparts of these networks with multiple transmit and receive antennas. The code for the Gaussian relay network can be obtained from any code for the discrete superposition network simply by pruning it. This lifting scheme establishes that the superposition model can indeed potentially serve as a strong surrogate for designing codes for Gaussian relay networks. We present similar results for the K x K Gaussian interference network, MIMO Gaussian interference networks, MIMO Gaussian relay networks, and multicast networks, with the constant gap depending additionally on the number of antennas in case of MIMO networks.
1005.0188
Generative and Latent Mean Map Kernels
cs.LG stat.ML
We introduce two kernels that extend the mean map, which embeds probability measures in Hilbert spaces. The generative mean map kernel (GMMK) is a smooth similarity measure between probabilistic models. The latent mean map kernel (LMMK) generalizes the non-iid formulation of Hilbert space embeddings of empirical distributions in order to incorporate latent variable models. When comparing certain classes of distributions, the GMMK exhibits beneficial regularization and generalization properties not shown for previous generative kernels. We present experiments comparing support vector machine performance using the GMMK and LMMK between hidden Markov models to the performance of other methods on discrete and continuous observation sequence data. The results suggest that, in many cases, the GMMK has generalization error competitive with or better than other methods.
1005.0198
Personnalisation de Syst\`emes OLAP Annot\'es
cs.DB
This paper deals with personalization of annotated OLAP systems. Data constellation is extended to support annotations and user preferences. Annotations reflect the decision-maker experience whereas user preferences enable users to focus on the most interesting data. User preferences allow annotated contextual recommendations helping the decision-maker during his/her multidimensional navigations.
1005.0201
Personnalisation de bases de donn\'ees multidimensionnelles
cs.DB
This paper deals with decision support systems resting on multidimensional modelling of data. Moreover, we intend to offer a set of concepts and mechanisms for personalized multidimensional database specifications. This personalization consists in associating weights to different components of a multidimensional schema. Personalization specifications are specified through the use of a language based on the principle of Event Condition Action. This personalisation determines multidimensional data display as well as their analyses (with the use of drilling or rotating operations).
1005.0202
Dictionary Optimization for Block-Sparse Representations
cs.IT math.IT
Recent work has demonstrated that using a carefully designed dictionary instead of a predefined one, can improve the sparsity in jointly representing a class of signals. This has motivated the derivation of learning methods for designing a dictionary which leads to the sparsest representation for a given set of signals. In some applications, the signals of interest can have further structure, so that they can be well approximated by a union of a small number of subspaces (e.g., face recognition and motion segmentation). This implies the existence of a dictionary which enables block-sparse representations of the input signals once its atoms are properly sorted into blocks. In this paper, we propose an algorithm for learning a block-sparsifying dictionary of a given set of signals. We do not require prior knowledge on the association of signals into groups (subspaces). Instead, we develop a method that automatically detects the underlying block structure. This is achieved by iteratively alternating between updating the block structure of the dictionary and updating the dictionary atoms to better fit the data. Our experiments show that for block-sparse data the proposed algorithm significantly improves the dictionary recovery ability and lowers the representation error compared to dictionary learning methods that do not employ block structure.
1005.0212
Construction graphique d'entrep\^ots et de magasins de donn\'ees
cs.DB
Nowadays, decisional systems have became a significant research topic in databases. Data warehouses and data marts are the main elements of such systems. This paper presents our decisional support system. We present graphical interfaces which help the administrator to build data warehouses and data marts. We present a data warehouse building interface based on an object-oriented conceptual model. This model allows the warehouse data historisation at three levels: attribute, class and environment. Also, we present a data mart building interface which allows warehouse data to be reorganised through a multidimensional object-oriented model.
1005.0213
Alg\`ebre OLAP et langage graphique
cs.DB
This article deals with OLAP systems based on multidimensional model. The conceptual model we provide, represents data through a constellation (multi-facts) composed of several multi-hierarchy dimensions. In this model, data are displayed through multidimensional tables. We define a query algebra handling these tables. This user oriented algebra is composed of a closure core of OLAP operators as soon as advanced operators dedicated to complex analysis. Finally, we specify a graphical OLAP language based on this algebra. This language facilitates analyses of decision makers.
1005.0214
Mod\'elisation et extraction de donn\'ees pour un entrep\^ot objet
cs.DB
This paper describes an object-oriented model for designing complex and time-variant data warehouse data. The main contribution is the warehouse class concept, which extends the class concept by temporal and archive filters as well as a mapping function. Filters allow the keeping of relevant data changes whereas the mapping function defines the warehouse class schema from a global data source schema. The approach take into account static properties as well as dynamic properties. The behaviour extraction is based on the use-matrix concept.
1005.0217
Analyse multigraduelle OLAP
cs.DB
Decisional systems are based on multidimensional databases improving OLAP analyses. The paper describes a new OLAP operator named "BLEND" to perform multigradual analyses. The operation transforms multidimensional structures during querying in order to analyse measures according to various granularity levels, which are reorganised into a single parameter. We study valid combinations of the operation in the context of strict hierarchies. First experimentations implement the operation in an R-OLAP framework showing the slight cost of this operation.
1005.0218
Contraintes pour mod\`ele et langage multidimensionnels
cs.DB
This paper defines a constraint-based model dedicated to multidimensional databases. The model we define represents data through a constellation of facts (subjects of analyse) associated to dimensions (axis of analyse), which are possibly shared. Each dimension is organised according to several hierarchies (views of analyse) integrating several levels of data granularity. In order to insure data consistency, we introduce 5 semantic constraints (exclusion, inclusion, partition, simultaneity, totality) which can be intra-dimension or inter-dimensions; the intra-dimension constraints allow the expression of constraints between hierarchies within a same dimension whereas the inter-dimensions constraints focus on hierarchies of distinct dimensions. We also study repercussions of these constraints on multidimensional manipulations and we provide extensions of the multidimensional operators.
1005.0219
Mod\'elisation et manipulation de donn\'ees historis\'ees et archiv\'ees dans un entrep\^ot orient\'e objet
cs.DB
This paper deals with temporal and archive object-oriented data warehouse modelling and querying. In a first step, we define a data model describing warehouses as central repositories of complex and temporal data extracted from one information source. The model is based on the concepts of warehouse object and environment. A warehouse object is composed of one current state, several past states (modelling value changes) and several archive states (summarising some value changes). An environment defines temporal parts in a warehouse schema according to a relevant granularity (attribute, class or graph). In a second step, we provide a query algebra dedicated to data warehouses. This algebra, which is based on common object algebras, integrates temporal operators and operators for querying object states. An other important contribution concerns dedicated operators allowing users to transform warehouse objects in temporal series as well as operators facilitating analytical treatments.
1005.0220
Elaboration d'entrep\^ots de donn\'ees complexes
cs.DB
In this paper, we study the data warehouse modelling used in decision support systems. We provide an object-oriented data warehouse model allowing data warehouse description as a central repository of relevant, complex and temporal data. Our model integrates three concepts such as warehouse object, environment and warehouse class. Each warehouse object is composed of one current state, several past states (modelling its detailed evolutions) and several archive states (modelling its evolutions within a summarised form). The environment concept defines temporal parts in the data warehouse schema with significant granularities (attribute, class, graph). Finally, we provide five functions aiming at defining the data warehouse structures and two functions allowing the warehouse class inheritance hierarchy organisation.
1005.0224
Towards Conceptual Multidimensional Design in Decision Support Systems
cs.DB
Multidimensional databases support efficiently on-line analytical processing (OLAP). In this paper, we depict a model dedicated to multidimensional databases. The approach we present designs decisional information through a constellation of facts and dimensions. Each dimension is possibly shared between several facts and it is organised according to multiple hierarchies. In addition, we define a comprehensive query algebra regrouping the more popular multidimensional operations in current commercial systems and research approaches. We introduce new operators dedicated to a constellation. Finally, we describe a prototype that allows managers to query constellations of facts, dimensions and multiple hierarchies.
1005.0267
Recovery of sparsest signals via $\ell^q$-minimization
cs.IT math.IT
In this paper, it is proved that every $s$-sparse vector ${\bf x}\in {\mathbb R}^n$ can be exactly recovered from the measurement vector ${\bf z}={\bf A} {\bf x}\in {\mathbb R}^m$ via some $\ell^q$-minimization with $0< q\le 1$, as soon as each $s$-sparse vector ${\bf x}\in {\mathbb R}^n$ is uniquely determined by the measurement ${\bf z}$.
1005.0268
Node-Context Network Clustering using PARAFAC Tensor Decomposition
cs.IR
We describe a clustering method for labeled link network (semantic graph) that can be used to group important nodes (highly connected nodes) with their relevant link's labels by using PARAFAC tensor decomposition. In this kind of network, the adjacency matrix can not be used to fully describe all information about the network structure. We have to expand the matrix into 3-way adjacency tensor, so that not only the information about to which nodes a node connects to but by which link's labels is also included. And by applying PARAFAC decomposition on this tensor, we get two lists, nodes and link's labels with scores attached to each node and labels, for each decomposition group. So clustering process to get the important nodes along with their relevant labels can be done simply by sorting the lists in decreasing order. To test the method, we construct labeled link network by using blog's dataset, where the blogs are the nodes and labeled links are the shared words among them. The similarity measures between the results and standard measures look promising, especially for two most important tasks, finding the most relevant words to blogs query and finding the most similar blogs to blogs query, about 0.87.
1005.0291
The Compound Multiple Access Channel with Partially Cooperating Encoders
cs.IT math.IT
The goal of this paper is to provide a rigorous information-theoretic analysis of subnetworks of interference networks. We prove two coding theorems for the compound multiple-access channel with an arbitrary number of channel states. The channel state information at the transmitters is such that each transmitter has a finite partition of the set of states and knows which element of the partition the actual state belongs to. The receiver may have arbitrary channel state information. The first coding theorem is for the case that both transmitters have a common message and that each has an additional common message. The second coding theorem is for the case where rate-constrained, but noiseless transmitter cooperation is possible. This cooperation may be used to exchange information about channel state information as well as the messages to be transmitted. The cooperation protocol used here generalizes Willems' conferencing. We show how this models base station cooperation in modern wireless cellular networks used for interference coordination and capacity enhancement. In particular, the coding theorem for the cooperative case shows how much cooperation is necessary in order to achieve maximal capacity in the network considered.
1005.0340
Statistical Learning in Automated Troubleshooting: Application to LTE Interference Mitigation
cs.LG
This paper presents a method for automated healing as part of off-line automated troubleshooting. The method combines statistical learning with constraint optimization. The automated healing aims at locally optimizing radio resource management (RRM) or system parameters of cells with poor performance in an iterative manner. The statistical learning processes the data using Logistic Regression (LR) to extract closed form (functional) relations between Key Performance Indicators (KPIs) and Radio Resource Management (RRM) parameters. These functional relations are then processed by an optimization engine which proposes new parameter values. The advantage of the proposed formulation is the small number of iterations required by the automated healing method to converge, making it suitable for off-line implementation. The proposed method is applied to heal an Inter-Cell Interference Coordination (ICIC) process in a 3G Long Term Evolution (LTE) network which is based on soft-frequency reuse scheme. Numerical simulations illustrate the benefits of the proposed approach.
1005.0375
Performance Analysis of Cognitive Radio Systems under QoS Constraints and Channel Uncertainty
cs.IT math.IT
In this paper, performance of cognitive transmission over time-selective flat fading channels is studied under quality of service (QoS) constraints and channel uncertainty. Cognitive secondary users (SUs) are assumed to initially perform channel sensing to detect the activities of the primary users, and then attempt to estimate the channel fading coefficients through training. Energy detection is employed for channel sensing, and different minimum mean-square-error (MMSE) estimation methods are considered for channel estimation. In both channel sensing and estimation, erroneous decisions can be made, and hence, channel uncertainty is not completely eliminated. In this setting, performance is studied and interactions between channel sensing and estimation are investigated. Following the channel sensing and estimation tasks, SUs engage in data transmission. Transmitter, being unaware of the channel fading coefficients, is assumed to send the data at fixed power and rate levels that depend on the channel sensing results. Under these assumptions, a state-transition model is constructed by considering the reliability of the transmissions, channel sensing decisions and their correctness, and the evolution of primary user activity which is modeled as a two-state Markov process. In the data transmission phase, an average power constraint on the secondary users is considered to limit the interference to the primary users, and statistical limitations on the buffer lengths are imposed to take into account the QoS constraints of the secondary traffic. The maximum throughput under these statistical QoS constraints is identified by finding the effective capacity of the cognitive radio channel. Numerical results are provided for the power and rate policies.
1005.0390
Machine Learning for Galaxy Morphology Classification
astro-ph.GA cs.LG
In this work, decision tree learning algorithms and fuzzy inferencing systems are applied for galaxy morphology classification. In particular, the CART, the C4.5, the Random Forest and fuzzy logic algorithms are studied and reliable classifiers are developed to distinguish between spiral galaxies, elliptical galaxies or star/unknown galactic objects. Morphology information for the training and testing datasets is obtained from the Galaxy Zoo project while the corresponding photometric and spectra parameters are downloaded from the SDSS DR7 catalogue.
1005.0404
Approximate Capacity of Gaussian Interference-Relay Networks with Weak Cross Links
cs.IT math.IT
In this paper we study a Gaussian relay-interference network, in which relay (helper) nodes are to facilitate competing information flows over a wireless network. We focus on a two-stage relay-interference network where there are weak cross-links, causing the networks to behave like a chain of Z Gaussian channels. For these Gaussian ZZ and ZS networks, we establish an approximate characterization of the rate region. The outer bounds to the capacity region are established using genie-aided techniques that yield bounds sharper than the traditional cut-set outer bounds. For the inner bound of the ZZ network, we propose a new interference management scheme, termed interference neutralization, which is implemented using structured lattice codes. This technique allows for over-the-air interference removal, without the transmitters having complete access the interfering signals. For both the ZZ and ZS networks, we establish a new network decomposition technique that (approximately) achieves the capacity region. We use insights gained from an exact characterization of the corresponding linear deterministic version of the problems, in order to establish the approximate characterization for Gaussian networks.
1005.0416
Incremental Sampling-based Algorithms for Optimal Motion Planning
cs.RO
During the last decade, incremental sampling-based motion planning algorithms, such as the Rapidly-exploring Random Trees (RRTs) have been shown to work well in practice and to possess theoretical guarantees such as probabilistic completeness. However, no theoretical bounds on the quality of the solution obtained by these algorithms have been established so far. The first contribution of this paper is a negative result: it is proven that, under mild technical conditions, the cost of the best path in the RRT converges almost surely to a non-optimal value. Second, a new algorithm is considered, called the Rapidly-exploring Random Graph (RRG), and it is shown that the cost of the best path in the RRG converges to the optimum almost surely. Third, a tree version of RRG is introduced, called the RRT$^*$ algorithm, which preserves the asymptotic optimality of RRG while maintaining a tree structure like RRT. The analysis of the new algorithms hinges on novel connections between sampling-based motion planning algorithms and the theory of random geometric graphs. In terms of computational complexity, it is shown that the number of simple operations required by both the RRG and RRT$^*$ algorithms is asymptotically within a constant factor of that required by RRT.
1005.0419
Capacity-Equivocation Region of the Gaussian MIMO Wiretap Channel
cs.IT cs.CR math.IT
We study the Gaussian multiple-input multiple-output (MIMO) wiretap channel, which consists of a transmitter, a legitimate user, and an eavesdropper. In this channel, the transmitter sends a common message to both the legitimate user and the eavesdropper. In addition to this common message, the legitimate user receives a private message, which is desired to be kept hidden as much as possible from the eavesdropper. We obtain the entire capacity-equivocation region of the Gaussian MIMO wiretap channel. This region contains all achievable common message, private message, and private message's equivocation (secrecy) rates. In particular, we show the sufficiency of jointly Gaussian auxiliary random variables and channel input to evaluate the existing single-letter description of the capacity-equivocation region due to Csiszar-Korner.
1005.0426
Security in Distributed Storage Systems by Communicating a Logarithmic Number of Bits
cs.CR cs.IT math.IT
We investigate the problem of maintaining an encoded distributed storage system when some nodes contain adversarial errors. Using the error-correction capabilities that are built into the existing redundancy of the system, we propose a simple linear hashing scheme to detect errors in the storage nodes. Our main result is that for storing a data object of total size $\size$ using an $(n,k)$ MDS code over a finite field $\F_q$, up to $t_1=\lfloor(n-k)/2\rfloor$ errors can be detected, with probability of failure smaller than $1/ \size$, by communicating only $O(n(n-k)\log \size)$ bits to a trusted verifier. Our result constructs small projections of the data that preserve the errors with high probability and builds on a pseudorandom generator that fools linear functions. The transmission rate achieved by our scheme is asymptotically equal to the min-cut capacity between the source and any receiver.
1005.0437
A Unifying View of Multiple Kernel Learning
stat.ML cs.LG
Recent research on multiple kernel learning has lead to a number of approaches for combining kernels in regularized risk minimization. The proposed approaches include different formulations of objectives and varying regularization strategies. In this paper we present a unifying general optimization criterion for multiple kernel learning and show how existing formulations are subsumed as special cases. We also derive the criterion's dual representation, which is suitable for general smooth optimization algorithms. Finally, we evaluate multiple kernel learning in this framework analytically using a Rademacher complexity bound on the generalization error and empirically in a set of experiments.
1005.0498
Classes of lower bounds on outage error probability and MSE in Bayesian parameter estimation
cs.IT math.IT
In this paper, new classes of lower bounds on the outage error probability and on the mean-square-error (MSE) in Bayesian parameter estimation are proposed. The minima of the h-outage error probability and the MSE are obtained by the generalized maximum a-posteriori probability and the minimum MSE (MMSE) estimators, respectively. However, computation of these estimators and their corresponding performance is usually not tractable and thus, lower bounds on these terms can be very useful for performance analysis. The proposed class of lower bounds on the outage error probability is derived using Holder's inequality. This class is utilized to derive a new class of Bayesian MSE bounds. It is shown that for unimodal symmetric conditional probability density functions (pdf) the tightest probability of outage error lower bound in the proposed class attains the minimum probability of outage error and the tightest MSE bound coincides with the MMSE performance. In addition, it is shown that the proposed MSE bounds are always tighter than the Ziv-Zakai lower bound (ZZLB). The proposed bounds are compared with other existing performance lower bounds via some examples.
1005.0527
Detecting the Most Unusual Part of Two and Three-dimensional Digital Images
physics.data-an cs.CV physics.med-ph
The purpose of this paper is to introduce an algorithm that can detect the most unusual part of a digital image in probabilistic setting. The most unusual part of a given shape is defined as a part of the image that has the maximal distance to all non intersecting shapes with the same form. The method is tested on two and three-dimensional images and has shown very good results without any predefined model. A version of the method independent of the contrast of the image is considered and is found to be useful for finding the most unusual part (and the most similar part) of the image conditioned on given image. The results can be used to scan large image databases, as for example medical databases.
1005.0530
Feature Selection with Conjunctions of Decision Stumps and Learning from Microarray Data
cs.LG cs.AI stat.ML
One of the objectives of designing feature selection learning algorithms is to obtain classifiers that depend on a small number of attributes and have verifiable future performance guarantees. There are few, if any, approaches that successfully address the two goals simultaneously. Performance guarantees become crucial for tasks such as microarray data analysis due to very small sample sizes resulting in limited empirical evaluation. To the best of our knowledge, such algorithms that give theoretical bounds on the future performance have not been proposed so far in the context of the classification of gene expression data. In this work, we investigate the premise of learning a conjunction (or disjunction) of decision stumps in Occam's Razor, Sample Compression, and PAC-Bayes learning settings for identifying a small subset of attributes that can be used to perform reliable classification tasks. We apply the proposed approaches for gene identification from DNA microarray data and compare our results to those of well known successful approaches proposed for the task. We show that our algorithm not only finds hypotheses with much smaller number of genes while giving competitive classification accuracy but also have tight risk guarantees on future performance unlike other approaches. The proposed approaches are general and extensible in terms of both designing novel algorithms and application to other domains.
1005.0545
Capacity of a Class of Broadcast Relay Channels
cs.IT math.IT
Consider the broadcast relay channel (BRC) which consists of a source sending information over a two user broadcast channel in presence of two relay nodes that help the transmission to the destinations. Clearly, this network with five nodes involves all the problems encountered in relay and broadcast channels. New inner bounds on the capacity region of this class of channels are derived. These results can be seen as a generalization and hence unification of previous work in this topic. Our bounds are based on the idea of recombination of message bits and various effective coding strategies for relay and broadcast channels. Capacity result is obtained for the semi-degraded BRC-CR, where one relay channel is degraded while the other one is reversely degraded. An inner and upper bound is also presented for the degraded BRC with common relay (BRC-CR), where both the relay and broadcast channel are degraded which is the capacity for the Gaussian case. Application of these results arise in the context of opportunistic cooperation of cellular networks.
1005.0605
An approach to visualize the course of solving of a research task in humans
cs.AI
A technique to study the dynamics of solving of a research task is suggested. The research task was based on specially developed software Right- Wrong Responder (RWR), with the participants having to reveal the response logic of the program. The participants interacted with the program in the form of a semi-binary dialogue, which implies the feedback responses of only two kinds - "right" or "wrong". The technique has been applied to a small pilot group of volunteer participants. Some of them have successfully solved the task (solvers) and some have not (non-solvers). In the beginning of the work, the solvers did more wrong moves than non-solvers, and they did less wrong moves closer to the finish of the work. A phase portrait of the work both in solvers and non-solvers showed definite cycles that may correspond to sequences of partially true hypotheses that may be formulated by the participants during the solving of the task.