id
stringlengths
9
16
title
stringlengths
4
278
categories
stringlengths
5
104
abstract
stringlengths
6
4.09k
1210.3741
Online computation of sparse representations of time varying stimuli using a biologically motivated neural network
q-bio.NC cs.NE
Natural stimuli are highly redundant, possessing significant spatial and temporal correlations. While sparse coding has been proposed as an efficient strategy employed by neural systems to encode sensory stimuli, the underlying mechanisms are still not well understood. Most previous approaches model the neural dynamics by the sparse representation dictionary itself and compute the representation coefficients offline. In reality, faced with the challenge of constantly changing stimuli, neurons must compute the sparse representations dynamically in an online fashion. Here, we describe a leaky linearized Bregman iteration (LLBI) algorithm which computes the time varying sparse representations using a biologically motivated network of leaky rectifying neurons. Compared to previous attempt of dynamic sparse coding, LLBI exploits the temporal correlation of stimuli and demonstrate better performance both in representation error and the smoothness of temporal evolution of sparse coefficients.
1210.3769
Analysis of Blocking Probability in a Relay-based Cellular OFDMA Network
cs.IT cs.NI math.IT
Relay deployment in Orthogonal Frequency Division Multiple Access (OFDMA) based cellular networks helps in coverage extension and or capacity improvement. In OFDMA system, each user requires different number of subcarriers to meet its rate requirement. This resource requirement depends on the Signal to Interference Ratio (SIR) experienced by a user. Traditional methods to compute blocking probability cannot be used in relay based cellular OFDMA networks. In this paper, we present an approach to compute the blocking probability of such networks. We determine an expression of the probability distribution of the users resource requirement based on its experienced SIR and then classify the users into various classes depending upon their subcarrier requirement. We consider the system to be a multidimensional system with different classes and evaluate the blocking probability of system using the multi-dimensional Erlang loss formulas.
1210.3812
A Unified Analytical Design Method of Standard Controllers using Inversion Formulae
cs.SY math.OC
The aim of this paper is to present a comprehensive range of design techniques for the synthesis of the standard compensators (Lead and Lag networks as well as PID controllers) that in the last twenty years have proved to be of great educational value in a vast number of undergraduate and postgraduate courses in Control throughout Italy, but that to-date remain mostly confined within this country. These techniques hinge upon a set of simple closed-form formulae for the computation of the parameters of the controller as functions of the typical specifications introduced in Control courses, i.e., the steady-state performance, the stability margins and the crossover frequencies.
1210.3819
On Precoding for Constant K-User MIMO Gaussian Interference Channel with Finite Constellation Inputs
cs.IT math.IT
This paper considers linear precoding for constant channel-coefficient $K$-User MIMO Gaussian Interference Channel (MIMO GIC) where each transmitter-$i$ (Tx-$i$), requires to send $d_i$ independent complex symbols per channel use that take values from fixed finite constellations with uniform distribution, to receiver-$i$ (Rx-$i$) for $i=1,2,\cdots,K$. We define the maximum rate achieved by Tx-$i$ using any linear precoder, when the interference channel-coefficients are zero, as the signal to noise ratio (SNR) tends to infinity to be the Constellation Constrained Saturation Capacity (CCSC) for Tx-$i$. We derive a high SNR approximation for the rate achieved by Tx-$i$ when interference is treated as noise and this rate is given by the mutual information between Tx-$i$ and Rx-$i$, denoted as $I[X_i;Y_i]$. A set of necessary and sufficient conditions on the precoders under which $I[X_i;Y_i]$ tends to CCSC for Tx-$i$ is derived. Interestingly, the precoders designed for interference alignment (IA) satisfy these necessary and sufficient conditions. Further, we propose gradient-ascent based algorithms to optimize the sum-rate achieved by precoding with finite constellation inputs and treating interference as noise. Simulation study using the proposed algorithms for a 3-user MIMO GIC with two antennas at each node with $d_i=1$ for all $i$, and with BPSK and QPSK inputs, show more than 0.1 bits/sec/Hz gain in the ergodic sum-rate over that yielded by precoders obtained from some known IA algorithms, at moderate SNRs.
1210.3832
Image Processing using Smooth Ordering of its Patches
cs.CV
We propose an image processing scheme based on reordering of its patches. For a given corrupted image, we extract all patches with overlaps, refer to these as coordinates in high-dimensional space, and order them such that they are chained in the "shortest possible path", essentially solving the traveling salesman problem. The obtained ordering applied to the corrupted image, implies a permutation of the image pixels to what should be a regular signal. This enables us to obtain good recovery of the clean image by applying relatively simple 1D smoothing operations (such as filtering or interpolation) to the reordered set of pixels. We explore the use of the proposed approach to image denoising and inpainting, and show promising results in both cases.
1210.3835
Exploiting Network Cooperation in Green Wireless Communication
cs.IT cs.PF math.IT
There is a growing interest in energy efficient or so-called "green" wireless communication to reduce the energy consumption in cellular networks. Since today's wireless terminals are typically equipped with multiple network access interfaces such as Bluetooth, Wi-Fi, and cellular networks, this paper investigates user terminals cooperating with each other in transmitting their data packets to the base station (BS), by exploiting the multiple network access interfaces, called inter-network cooperation. We also examine the conventional schemes without user cooperation and with intra-network cooperation for comparison. Given target outage probability and data rate requirements, we analyze the energy consumption of conventional schemes as compared to the proposed inter-network cooperation by taking into account both physical-layer channel impairments (including path loss, fading, and thermal noise) and upper-layer protocol overheads. It is shown that distances between different network entities (i.e., user terminals and BS) have a significant influence on the energy efficiency of proposed inter-network cooperation scheme. Specifically, when the cooperating users are close to BS or the users are far away from each other, the inter-network cooperation may consume more energy than conventional schemes without user cooperation or with intra-network cooperation. However, as the cooperating users move away from BS and the inter-user distance is not too large, the inter-network cooperation significantly reduces the energy consumption over conventional schemes.
1210.3853
Transceiver Design For SC-FDE Based MIMO Relay Systems
cs.IT math.IT
In this paper, we propose a joint transceiver design for single-carrier frequency-domain equalization (SC-FDE) based multiple-input multiple-output (MIMO) relay systems. To this end, we first derive the optimal minimum mean-squared error linear and decision-feedback frequency-domain equalization filters at the destination along with the corresponding error covariance matrices at the output of the equalizer. Subsequently, we formulate the source and relay precoding matrix design problem as the minimization of a family of Schur-convex and Schur-concave functions of the mean-squared errors at the output of the equalizer under separate power constraints for the source and the relay. By exploiting properties of the error covariance matrix and results from majorization theory, we derive the optimal structures of the source and relay precoding matrices, which allows us to transform the matrix optimization problem into a scalar power optimization problem. Adopting a high signal-to-noise ratio approximation for the objective function, we obtain the global optimal solution for the power allocation variables. Simulation results illustrate the excellent performance of the proposed system and its superiority compared to conventional orthogonal frequency-division multiplexing based MIMO relay systems.
1210.3865
Opinion Mining for Relating Subjective Expressions and Annual Earnings in US Financial Statements
cs.CL cs.AI cs.IR q-fin.GN
Financial statements contain quantitative information and manager's subjective evaluation of firm's financial status. Using information released in U.S. 10-K filings. Both qualitative and quantitative appraisals are crucial for quality financial decisions. To extract such opinioned statements from the reports, we built tagging models based on the conditional random field (CRF) techniques, considering a variety of combinations of linguistic factors including morphology, orthography, predicate-argument structure, syntax, and simple semantics. Our results show that the CRF models are reasonably effective to find opinion holders in experiments when we adopted the popular MPQA corpus for training and testing. The contribution of our paper is to identify opinion patterns in multiword expressions (MWEs) forms rather than in single word forms. We find that the managers of corporations attempt to use more optimistic words to obfuscate negative financial performance and to accentuate the positive financial performance. Our results also show that decreasing earnings were often accompanied by ambiguous and mild statements in the reporting year and that increasing earnings were stated in assertive and positive way.
1210.3906
Design of Multiple-Edge Protographs for QC LDPC Codes Avoiding Short Inevitable Cycles
cs.IT math.IT
There have been lots of efforts on the construction of quasi-cyclic (QC) low-density parity-check (LDPC) codes with large girth. However, most of them are focused on protographs with single edges and little research has been done for the construction of QC LDPC codes lifted from protographs with multiple edges. Compared to single-edge protographs, multiple-edge protographs have benefits such that QC LDPC codes lifted from them can potentially have larger minimum Hamming distance. In this paper, all subgraph patterns of multiple-edge protographs, which prevent QC LDPC codes from having large girth by inducing inevitable cycles, are fully investigated based on graph-theoretic approach. By using combinatorial designs, a systematic construction method of multiple-edge protographs is proposed for regular QC LDPC codes with girth at least 12 and also other method is proposed for regular QC LDPC codes with girth at least 14. A construction algorithm of QC LDPC codes by lifting multiple-edge protographs is proposed and it is shown that the resulting QC LDPC codes have larger upper bounds on the minimum Hamming distance than those lifted from single-edge protographs. Simulation results are provided to compare the performance of the proposed QC LDPC codes, the progressive edge-growth (PEG) LDPC codes, and the PEG QC LDPC codes.
1210.3921
Stein's density approach and information inequalities
math.PR cs.IT math.IT
We provide a new perspective on Stein's so-called density approach by introducing a new operator and characterizing class which are valid for a much wider family of probability distributions on the real line. We prove an elementary factorization property of this operator and propose a new Stein identity which we use to derive information inequalities in terms of what we call the \emph{generalized Fisher information distance}. We provide explicit bounds on the constants appearing in these inequalities for several important cases. We conclude with a comparison between our results and known results in the Gaussian case, hereby improving on several known inequalities from the literature.
1210.3926
Learning Attitudes and Attributes from Multi-Aspect Reviews
cs.CL cs.IR cs.LG
The majority of online reviews consist of plain-text feedback together with a single numeric score. However, there are multiple dimensions to products and opinions, and understanding the `aspects' that contribute to users' ratings may help us to better understand their individual preferences. For example, a user's impression of an audiobook presumably depends on aspects such as the story and the narrator, and knowing their opinions on these aspects may help us to recommend better products. In this paper, we build models for rating systems in which such dimensions are explicit, in the sense that users leave separate ratings for each aspect of a product. By introducing new corpora consisting of five million reviews, rated with between three and six aspects, we evaluate our models on three prediction tasks: First, we use our model to uncover which parts of a review discuss which of the rated aspects. Second, we use our model to summarize reviews, which for us means finding the sentences that best explain a user's rating. Finally, since aspect ratings are optional in many of the datasets we consider, we use our model to recover those ratings that are missing from a user's evaluation. Our model matches state-of-the-art approaches on existing small-scale datasets, while scaling to the real-world datasets we introduce. Moreover, our model is able to `disentangle' content and sentiment words: we automatically learn content words that are indicative of a particular aspect as well as the aspect-specific sentiment words that are indicative of a particular rating.
1210.3937
Introduction to the 28th International Conference on Logic Programming Special Issue
cs.PL cs.AI
We are proud to introduce this special issue of the Journal of Theory and Practice of Logic Programming (TPLP), dedicated to the full papers accepted for the 28th International Conference on Logic Programming (ICLP). The ICLP meetings started in Marseille in 1982 and since then constitute the main venue for presenting and discussing work in the area of logic programming.
1210.3946
Local optima networks and the performance of iterated local search
cs.AI
Local Optima Networks (LONs) have been recently proposed as an alternative model of combinatorial fitness landscapes. The model compresses the information given by the whole search space into a smaller mathematical object that is the graph having as vertices the local optima and as edges the possible weighted transitions between them. A new set of metrics can be derived from this model that capture the distribution and connectivity of the local optima in the underlying configuration space. This paper departs from the descriptive analysis of local optima networks, and actively studies the correlation between network features and the performance of a local search heuristic. The NK family of landscapes and the Iterated Local Search metaheuristic are considered. With a statistically-sound approach based on multiple linear regression, it is shown that some LONs' features strongly influence and can even partly predict the performance of a heuristic search algorithm. This study validates the expressive power of LONs as a model of combinatorial fitness landscapes.
1210.3953
Wireless Network-Coded Four-Way Relaying Using Latin Hyper-Cubes
cs.IT math.IT
This paper deals with physical layer network-coding for the four-way wireless relaying scenario where four nodes A, B, C and D wish to communicate their messages to all the other nodes with the help of the relay node R. The scheme given in the paper is based on the denoise-and-forward scheme proposed first by Popovski et al. Intending to minimize the number of channel uses, the protocol employs two phases: Multiple Access (MA) phase and Broadcast (BC) phase with each phase utilizing one channel use. This paper does the equivalent for the four-way relaying scenario as was done for the two-way relaying scenario by Koike-Akino et al., and for three-way relaying scenario in [3]. It is observed that adaptively changing the network coding map used at the relay according to the channel conditions greatly reduces the impact of multiple access interference which occurs at the relay during the MA phase. These network coding maps are so chosen so that they satisfy a requirement called exclusive law. We show that when the four users transmit points from the same M-PSK constellation, every such network coding map that satisfies the exclusive law can be represented by a 4-fold Latin Hyper-Cube of side M. The network code map used by the relay for the BC phase is explicitly obtained and is aimed at reducing the effect of interference at the MA stage.
1210.4006
The Perturbed Variation
cs.LG stat.ML
We introduce a new discrepancy score between two distributions that gives an indication on their similarity. While much research has been done to determine if two samples come from exactly the same distribution, much less research considered the problem of determining if two finite samples come from similar distributions. The new score gives an intuitive interpretation of similarity; it optimally perturbs the distributions so that they best fit each other. The score is defined between distributions, and can be efficiently estimated from samples. We provide convergence bounds of the estimated score, and develop hypothesis testing procedures that test if two data sets come from similar distributions. The statistical power of this procedures is presented in simulations. We also compare the score's capacity to detect similarity with that of other known measures on real data.
1210.4007
Extending modularity by capturing the similarity attraction feature in the null model
cs.SI physics.data-an physics.soc-ph
Modularity is a widely used measure for evaluating community structure in networks. The definition of modularity involves a comparison of within-community edges in the observed network and that number in an equivalent randomized network. This equivalent randomized network is called the null model, which serves as a reference. To make the comparison significant, the null model should characterize some features of the observed network. However, the null model in the original definition of modularity is unrealistically mixed, in the sense that any node can be linked to any other node without preference and only connectivity matters. Thus, it fails to be a good representation of real-world networks. A common feature of many real-world networks is "similarity attraction", i.e., edges tend to link to nodes that are similar to each other. We propose a null model that captures the similarity attraction feature. This null model enables us to create a framework for defining a family of Dist-Modularity adapted to various networks, including networks with additional information on nodes. We demonstrate that Dist-Modularity is useful in identifying communities at different scales.
1210.4008
Location-Based Events Detection on Micro-Blogs
cs.SI cs.IR physics.soc-ph
The increasing use of social networks generates enormous amounts of data that can be used for many types of analysis. Some of these data have temporal and geographical information, which can be used for comprehensive examination. In this paper, we propose a new method to analyze the massive volume of messages available in Twitter to identify places in the world where topics such as TV shows, climate change, disasters, and sports are emerging. The proposed method is based on a neural network that is used to detect outliers from a time series, which is built upon statistical data from tweets located on different political divisions (i.e., countries, cities). The outliers are used to identify topics within an abnormal behavior in Twitter. The effectiveness of our method is evaluated in an online environment indicating new findings on modeling local people's behavior from different places.
1210.4021
Local Optima Networks, Landscape Autocorrelation and Heuristic Search Performance
cs.AI cs.NE
Recent developments in fitness landscape analysis include the study of Local Optima Networks (LON) and applications of the Elementary Landscapes theory. This paper represents a first step at combining these two tools to explore their ability to forecast the performance of search algorithms. We base our analysis on the Quadratic Assignment Problem (QAP) and conduct a large statistical study over 600 generated instances of different types. Our results reveal interesting links between the network measures, the autocorrelation measures and the performance of heuristic search algorithms.
1210.4081
Getting Feasible Variable Estimates From Infeasible Ones: MRF Local Polytope Study
cs.NA cs.CV cs.DS cs.LG math.OC
This paper proposes a method for construction of approximate feasible primal solutions from dual ones for large-scale optimization problems possessing certain separability properties. Whereas infeasible primal estimates can typically be produced from (sub-)gradients of the dual function, it is often not easy to project them to the primal feasible set, since the projection itself has a complexity comparable to the complexity of the initial problem. We propose an alternative efficient method to obtain feasibility and show that its properties influencing the convergence to the optimum are similar to the properties of the Euclidean projection. We apply our method to the local polytope relaxation of inference problems for Markov Random Fields and demonstrate its superiority over existing methods.
1210.4130
Relational Theories with Null Values and Non-Herbrand Stable Models
cs.LO cs.AI cs.DB
Generalized relational theories with null values in the sense of Reiter are first-order theories that provide a semantics for relational databases with incomplete information. In this paper we show that any such theory can be turned into an equivalent logic program, so that models of the theory can be generated using computational methods of answer set programming. As a step towards this goal, we develop a general method for calculating stable models under the domain closure assumption but without the unique name assumption.
1210.4145
A Biologically Realistic Model of Saccadic Eye Control with Probabilistic Population Codes
cs.NE q-bio.NC
The posterior parietal cortex is believed to direct eye movements, especially in regards to target tracking tasks, and a number of debates exist over the precise nature of the computations performed by the parietal cortex, with each side supported by different sets of biological evidence. In this paper I will present my model which navigates a course between some of these debates, towards the end of presenting a model which can explain some of the competing interpretations among the data sets. In particular, rather than assuming that proprioception or efference copies form the key source of information for computing eye position information, I use a biological plausible implementation of a Kalman filter to optimally combine the two signals, and a simple gain control mechanism in order to accommodate the latency of the proprioceptive signal. Fitting within the Bayesian brain hypothesis, the result is a Bayes optimal solution to the eye control problem, with a range of data supporting claims of biological plausibility.
1210.4184
The Kernel Pitman-Yor Process
cs.LG cs.AI stat.ML
In this work, we propose the kernel Pitman-Yor process (KPYP) for nonparametric clustering of data with general spatial or temporal interdependencies. The KPYP is constructed by first introducing an infinite sequence of random locations. Then, based on the stick-breaking construction of the Pitman-Yor process, we define a predictor-dependent random probability measure by considering that the discount hyperparameters of the Beta-distributed random weights (stick variables) of the process are not uniform among the weights, but controlled by a kernel function expressing the proximity between the location assigned to each weight and the given predictors.
1210.4211
Profit Maximization over Social Networks
cs.SI cs.GT physics.soc-ph
Influence maximization is the problem of finding a set of influential users in a social network such that the expected spread of influence under a certain propagation model is maximized. Much of the previous work has neglected the important distinction between social influence and actual product adoption. However, as recognized in the management science literature, an individual who gets influenced by social acquaintances may not necessarily adopt a product (or technology), due, e.g., to monetary concerns. In this work, we distinguish between influence and adoption by explicitly modeling the states of being influenced and of adopting a product. We extend the classical Linear Threshold (LT) model to incorporate prices and valuations, and factor them into users' decision-making process of adopting a product. We show that the expected profit function under our proposed model maintains submodularity under certain conditions, but no longer exhibits monotonicity, unlike the expected influence spread function. To maximize the expected profit under our extended LT model, we employ an unbudgeted greedy framework to propose three profit maximization algorithms. The results of our detailed experimental study on three real-world datasets demonstrate that of the three algorithms, \textsf{PAGE}, which assigns prices dynamically based on the profit potential of each candidate seed, has the best performance both in the expected profit achieved and in running time.
1210.4231
An example illustrating the imprecision of the efficient approach for diagnosis of Petri nets via integer linear programming
cs.SY cs.AI
This document demonstrates that the efficient approach for diagnosis of Petri nets via integer linear programming may be unable to detect a fault even if the system is diagnosable.
1210.4235
Node Classification in Networks of Stochastic Evidence Accumulators
cs.SY math.OC
This paper considers a network of stochastic evidence accumulators, each represented by a drift-diffusion model accruing evidence towards a decision in continuous time by observing a noisy signal and by exchanging information with other units according to a fixed communication graph. We bring into focus the relationship between the location of each unit in the communication graph and its certainty as measured by the inverse of the variance of its state. We show that node classification according to degree distributions or geodesic distances cannot faithfully capture node ranking in terms of certainty. Instead, all possible paths connecting each unit with the rest in the network must be incorporated. We make this precise by proving that node classification according to information centrality provides a rank ordering with respect to node certainty, thereby affording a direct interpretation of the certainty level of each unit in terms of the structural properties of the underlying communication graph.
1210.4243
Outage Probability Analysis of Dual Hop Relay Networks in Presence of Interference
cs.IT math.IT
Cooperative relaying improves the performance of wireless networks by forming a network of multiple independent virtual sources transmitting the same information as the source node. However, interference induced in the network reduces the performance of cooperative communications. In this work the statistical properties, the cumulative distribution function (CDF) and the probability density function (PDF) for a basic dual hop cooperative relay network with an arbitrary number of interferers over Rayleigh fading channels are derived. Two system models are considered: in the first system model, the interferers are only at the relay node; and in the second system model, interferers are both at the relay and the destination. This work is further extended to Nakagami-m faded interfering channels. Simulation results are presented on outage probability performance to verify the theoretical analysis.
1210.4246
A Latent Parameter Node-Centric Model for Spatial Networks
cs.SI physics.soc-ph
Spatial networks, in which nodes and edges are embedded in space, play a vital role in the study of complex systems. For example, many social networks attach geo-location information to each user, allowing the study of not only topological interactions between users, but spatial interactions as well. The defining property of spatial networks is that edge distances are associated with a cost, which may subtly influence the topology of the network. However, the cost function over distance is rarely known, thus developing a model of connections in spatial networks is a difficult task. In this paper, we introduce a novel model for capturing the interaction between spatial effects and network structure. Our approach represents a unique combination of ideas from latent variable statistical models and spatial network modeling. In contrast to previous work, we view the ability to form long/short-distance connections to be dependent on the individual nodes involved. For example, a node's specific surroundings (e.g. network structure and node density) may make it more likely to form a long distance link than other nodes with the same degree. To capture this information, we attach a latent variable to each node which represents a node's spatial reach. These variables are inferred from the network structure using a Markov Chain Monte Carlo algorithm. We experimentally evaluate our proposed model on 4 different types of real-world spatial networks (e.g. transportation, biological, infrastructure, and social). We apply our model to the task of link prediction and achieve up to a 35% improvement over previous approaches in terms of the area under the ROC curve. Additionally, we show that our model is particularly helpful for predicting links between nodes with low degrees. In these cases, we see much larger improvements over previous models.
1210.4247
Deterministic Selection of Phase Sequences in Low Complexity SLM Scheme
cs.IT math.IT
Selected mapping (SLM) is a suitable scheme, which can solve the peak-to-average power ratio (PAPR) problem. Recently, many researchers have concentrated on reducing the computational complexity of the SLM schemes. One of the low complexity SLM schemes is the Class III SLM scheme which uses only one inverse fast fourier transform (IFFT) operation for generating one orthogonal frequency division multiplexing (OFDM) signal sequence. By selecting rotations and cyclic shifts randomly, it can generate $N^3$ alternative OFDM signal sequences, where $N$ is the FFT size. But this selection can not guarantee the optimal PAPR reduction performances. Therefore, in this paper, we propose a simple deterministic cyclic shifts selection method which is optimal in case of having low variance of correlation coefficient between two alternative OFDM signal sequences. And we show that cyclic shifts are highly dependent on the PAPR reduction performance than rotations. For small FFT size and the number of alternative signal sequences is close to $N/8$, simulation results show that the proposed scheme can achieve better PAPR reduction performance than the Class III SLM scheme.
1210.4251
Performance Analysis Cluster and GPU Computing Environment on Molecular Dynamic Simulation of BRV-1 and REM2 with GROMACS
cs.DC cs.CE q-bio.BM
One of application that needs high performance computing resources is molecular d ynamic. There is some software available that perform molecular dynamic, one of these is a well known GROMACS. Our previous experiment simulating molecular dynamics of Indonesian grown herbal compounds show sufficient speed up on 32 n odes Cluster computing environment. In order to obtain a reliable simulation, one usually needs to run the experiment on the scale of hundred nodes. But this is expensive to develop and maintain. Since the invention of Graphical Processing Units that is also useful for general programming, many applications have been developed to run on this. This paper reports our experiments that evaluate the performance of GROMACS that runs on two different environment, Cluster computing resources and GPU based PCs. We run the experiment on BRV-1 and REM2 compounds. Four different GPUs are installed on the same type of PCs of quad cores; they are Gefore GTS 250, GTX 465, GTX 470 and Quadro 4000. We build a cluster of 16 nodes based on these four quad cores PCs. The preliminary experiment shows that those run on GTX 470 is the best among the other type of GPUs and as well as the cluster computing resource. A speed up around 11 and 12 is gained, while the cost of computer with GPU is only about 25 percent that of Cluster we built.
1210.4276
Semi-Supervised Classification Through the Bag-of-Paths Group Betweenness
stat.ML cs.LG
This paper introduces a novel, well-founded, betweenness measure, called the Bag-of-Paths (BoP) betweenness, as well as its extension, the BoP group betweenness, to tackle semisupervised classification problems on weighted directed graphs. The objective of semi-supervised classification is to assign a label to unlabeled nodes using the whole topology of the graph and the labeled nodes at our disposal. The BoP betweenness relies on a bag-of-paths framework assigning a Boltzmann distribution on the set of all possible paths through the network such that long (high-cost) paths have a low probability of being picked from the bag, while short (low-cost) paths have a high probability of being picked. Within that context, the BoP betweenness of node j is defined as the sum of the a posteriori probabilities that node j lies in-between two arbitrary nodes i, k, when picking a path starting in i and ending in k. Intuitively, a node typically receives a high betweenness if it has a large probability of appearing on paths connecting two arbitrary nodes of the network. This quantity can be computed in closed form by inverting a n x n matrix where n is the number of nodes. For the group betweenness, the paths are constrained to start and end in nodes within the same class, therefore defining a group betweenness for each class. Unlabeled nodes are then classified according to the class showing the highest group betweenness. Experiments on various real-world data sets show that BoP group betweenness outperforms all the tested state of-the-art methods. The benefit of the BoP betweenness is particularly noticeable when only a few labeled nodes are available.
1210.4277
Improving Smoothed l0 Norm in Compressive Sensing Using Adaptive Parameter Selection
cs.IT math.IT
Signal reconstruction in compressive sensing involves finding a sparse solution that satisfies a set of linear constraints. Several approaches to this problem have been considered in existing reconstruction algorithms. They each provide a trade-off between reconstruction capabilities and required computation time. In an attempt to push the limits for this trade-off, we consider a smoothed l0 norm (SL0) algorithm in a noiseless setup. We argue that using a set of carefully chosen parameters in our proposed adaptive SL0 algorithm may result in significantly better reconstruction capabilities in terms of phase transition while retaining the same required computation time as existing SL0 algorithms. A large set of simulations further support this claim. Simulations even reveal that the theoretical l1 curve may be surpassed in major parts of the phase space.
1210.4290
A Fast Iterative Algorithm for Recovery of Sparse Signals from One-Bit Quantized Measurements
cs.IT math.IT
This paper considers the problem of reconstructing sparse or compressible signals from one-bit quantized measurements. We study a new method that uses a log-sum penalty function, also referred to as the Gaussian entropy, for sparse signal recovery. Also, in the proposed method, sigmoid functions are introduced to quantify the consistency between the acquired one-bit quantized data and the reconstructed measurements. A fast iterative algorithm is developed by iteratively minimizing a convex surrogate function that bounds the original objective function, which leads to an iterative reweighted process that alternates between estimating the sparse signal and refining the weights of the surrogate function. Connections between the proposed algorithm and other existing methods are discussed. Numerical results are provided to illustrate the effectiveness of the proposed algorithm.
1210.4293
Communications with decode-and-forward relays in mesh networks
cs.IT math.IT
We consider mesh networks composed of groups of relaying nodes which operate in decode-and-forward mode, where each node from a group relays information to all the nodes in the next group. We study these networks in two setups, one where the nodes have complete channel state information from the nodes that transmit to them, and another when they only have the statistics of the channel. We derive recursive expressions for the probabilities of errors of the nodes and present several implementations of detectors used in these networks. We compare the mesh networks with multihop networks, the latter being formed by a set of parallel sections of multiple relaying nodes. We demonstrate with numerous simulations that there are significant improvements in performance of mesh over multihop networks in various scenarios.
1210.4301
Reputation Aggregation in Peer-to-Peer Network Using Differential Gossip Algorithm
cs.NI cs.SI
Reputation aggregation in peer to peer networks is generally a very time and resource consuming process. Moreover, most of the methods consider that a node will have same reputation with all the nodes in the network, which is not true. This paper proposes a reputation aggregation algorithm that uses a variant of gossip algorithm called differential gossip. In this paper, estimate of reputation is considered to be having two parts, one common component which is same with every node, and the other one is information received from immediate neighbours based on the neighbours' direct interaction with the node. The differential gossip is fast and requires less amount of resources. This mechanism allows computation of independent reputation value by a node, of every other node in the network, for each node. The differential gossip trust has been investigated for a power law network formed using preferential attachment \emph{(PA)} Model. The reputation computed using differential gossip trust shows good amount of immunity to the collusion. We have verified the performance of the algorithm on the power law networks of different sizes ranging from 100 nodes to 50,000 nodes.
1210.4329
Impact of Scheduling in the Return-Link of Multi-Beam Satellite MIMO Systems
cs.IT math.IT
The utilization of universal frequency reuse in multi-beam satellite systems introduces a non-negligible level of co-channel interference (CCI), which in turn penalizes the quality of service experienced by users. Taking this as starting point, the paper focuses on resource management performed by the gateway (hub) on the return-link, with particular emphasis on a scheduling algorithm based on bipartite graph approach. The study gives important insights into the achievable per-user rate and the role played by the number of users and spot beams considered for scheduling. More interestingly, it is shown that a free-slot assignment strategy helps to exploit the available satellite resources, thus guaranteeing a max-min rate requirement to users. Remarks about the trade-off between efficiency-loss and performance increase are finally drawn at the end of the paper.
1210.4347
Hilbert Space Embedding for Dirichlet Process Mixtures
stat.ML cs.LG
This paper proposes a Hilbert space embedding for Dirichlet Process mixture models via a stick-breaking construction of Sethuraman. Although Bayesian nonparametrics offers a powerful approach to construct a prior that avoids the need to specify the model size/complexity explicitly, an exact inference is often intractable. On the other hand, frequentist approaches such as kernel machines, which suffer from the model selection/comparison problems, often benefit from efficient learning algorithms. This paper discusses the possibility to combine the best of both worlds by using the Dirichlet Process mixture model as a case study.
1210.4377
Order statistics of observed network degrees
stat.ME cs.SI physics.soc-ph
This article discusses the properties of extremes of degree sequences calculated from network data. We introduce the notion of a normalized degree, in order to permit a comparison of degree sequences between networks with differing numbers of nodes. We model each normalized degree as a bounded continuous random variable, and determine the properties of the ordered k-maxima and minima of the normalized network degrees when they comprise a random sample from a Beta distribution. In this setting, their means and variances take a simplified form given by their ordering, and we discuss the relation of these quantities to other prescribed decays such as power laws. We verify the derived properties from simulated sets of normalized degrees, and discuss possible extensions to more flexible classes of distributions.
1210.4383
Distributed Formation of Balanced and Bistochastic Weighted Diagraphs in Multi-Agent Systems
cs.MA
Consensus strategies find a variety of applications in distributed coordination and decision making in multi-agent systems. In particular, average consensus plays a key role in a number of applications and is closely associated with two classes of digraphs, weight-balanced (for continuous-time systems) and bistochastic (for discrete-time systems). A weighted digraph is called balanced if, for each node, the sum of the weights of the edges outgoing from that node is equal to the sum of the weights of the edges incoming to that node. In addition, a weight-balanced digraph is bistochastic if all weights are nonnegative and, for each node, the sum of weights of edges incoming to that node and the sum of the weights of edges out-going from that node is unity; this implies that the corresponding weight matrix is column and row stochastic (i.e., doubly stochastic). We propose two distributed algorithms: one solves the weight-balance problem and the other solves the bistochastic matrix formation problem for a distributed system whose components (nodes) can exchange information via interconnection links (edges) that form an arbitrary, possibly directed, strongly connected communication topology (digraph). Both distributed algorithms achieve their goals asymptotically and operate iteratively by having each node adapt the (nonnegative) weights on its outgoing edges based on the weights of its incoming links (i.e., based on purely local information). We also provide examples to illustrate the operation, performance, and potential advantages of the proposed algorithms.
1210.4405
Semantic integration and analysis of clinical data
cs.DB
There is a growing need to semantically process and integrate clinical data from different sources for Clinical Data Management and Clinical Decision Support in the healthcare IT industry. In the clinical practice domain, the semantic gap between clinical information systems and domain ontologies is quite often difficult to bridge in one step. In this paper, we report our experience in using a two-step formalization approach to formalize clinical data, i.e. from database schemas to local formalisms and from local formalisms to domain (unifying) formalisms. We use N3 rules to explicitly and formally state the mapping from local ontologies to domain ontologies. The resulting data expressed in domain formalisms can be integrated and analyzed, though originating from very distinct sources. Practices of applying the two-step approach in the infectious disorders and cancer domains are introduced.
1210.4416
A Direct Proof of a Theorem Concerning Singular Hamiltonian Systems
cs.SY
This technical report presents a direct proof of Theorem~1 in [1] and some consequences that also account for (20) in [1]. This direct proof exploits a state space change of basis which replaces the coupled difference equations (10) in [1] with two equivalent difference equations which, instead, are decoupled.
1210.4459
Efficient Computation of Pareto Optimal Beamforming Vectors for the MISO Interference Channel with Successive Interference Cancellation
cs.IT math.IT
We study the two-user multiple-input single-output (MISO) Gaussian interference channel where the transmitters have perfect channel state information and employ single-stream beamforming. The receivers are capable of performing successive interference cancellation, so when the interfering signal is strong enough, it can be decoded, treating the desired signal as noise, and subtracted from the received signal, before the desired signal is decoded. We propose efficient methods to compute the Pareto-optimal rate points and corresponding beamforming vector pairs, by maximizing the rate of one link given the rate of the other link. We do so by splitting the original problem into four subproblems corresponding to the combinations of the receivers' decoding strategies - either decode the interference or treat it as additive noise. We utilize recently proposed parameterizations of the optimal beamforming vectors to equivalently reformulate each subproblem as a quasi-concave problem, which we solve very efficiently either analytically or via scalar numerical optimization. The computational complexity of the proposed methods is several orders-of-magnitude less than the complexity of the state-of-the-art methods. We use the proposed methods to illustrate the effect of the strength and spatial correlation of the channels on the shape of the rate region.
1210.4460
Fast SVM-based Feature Elimination Utilizing Data Radius, Hard-Margin, Soft-Margin
stat.ML cs.LG
Margin maximization in the hard-margin sense, proposed as feature elimination criterion by the MFE-LO method, is combined here with data radius utilization to further aim to lower generalization error, as several published bounds and bound-related formulations pertaining to lowering misclassification risk (or error) pertain to radius e.g. product of squared radius and weight vector squared norm. Additionally, we propose additional novel feature elimination criteria that, while instead being in the soft-margin sense, too can utilize data radius, utilizing previously published bound-related formulations for approaching radius for the soft-margin sense, whereby e.g. a focus was on the principle stated therein as "finding a bound whose minima are in a region with small leave-one-out values may be more important than its tightness". These additional criteria we propose combine radius utilization with a novel and computationally low-cost soft-margin light classifier retraining approach we devise named QP1; QP1 is the soft-margin alternative to the hard-margin LO. We correct an error in the MFE-LO description, find MFE-LO achieves the highest generalization accuracy among the previously published margin-based feature elimination (MFE) methods, discuss some limitations of MFE-LO, and find our novel methods herein outperform MFE-LO, attain lower test set classification error rate. On several datasets that each both have a large number of features and fall into the `large features few samples' dataset category, and on datasets with lower (low-to-intermediate) number of features, our novel methods give promising results. Especially, among our methods the tunable ones, that do not employ (the non-tunable) LO approach, can be tuned more aggressively in the future than herein, to aim to demonstrate for them even higher performance than herein.
1210.4469
A Rule-based Model of a Hypothetical Zombie Outbreak: Insights on the role of emotional factors during behavioral adaptation of an artificial population
q-bio.PE cs.MA cs.SI physics.soc-ph
Models of infectious diseases have been developed since the first half of the twentieth century. Most models haven't considered the role that emotional factors of the individual may play on the population's behavioral adaptation during the spread of a pandemic disease. Considering that local interactions among individuals generate patterns that -at a large scale- govern the action of masses, we have studied the behavioral adaptation of a population induced by the spread of an infectious disease. Therefore, we have developed a rule-based model of a hypothetical zombie outbreak, written in Kappa language, and simulated using Guillespie's stochastic approach. Our study addresses the specificity and heterogeneity of the system at the individual level, a highly desirable characteristic, mostly overlooked in classic epidemic models. Together with the basic elements of a typical epidemiological model, our model includes an individual representation of the disease progression and the traveling of agents among cities being affected. It also introduces an approximation to measure the effect of panic in the population as a function of the individual situational awareness. In addition, the effect of two possible countermeasures to overcome the zombie threat is considered: the availability of medical treatment and the deployment of special armed forces. However, due to the special characteristics of this hypothetical infectious disease, even using exaggerated numbers of countermeasures, only a small percentage of the population can be saved at the end of the simulations. As expected from a rule-based model approach, the global dynamics of our model resulted primarily governed by the mechanistic description of local interactions occurring at the individual level. As a whole, people's situational awareness resulted essential to modulate the inner dynamics of the system.
1210.4481
Epitome for Automatic Image Colorization
cs.CV cs.LG cs.MM
Image colorization adds color to grayscale images. It not only increases the visual appeal of grayscale images, but also enriches the information contained in scientific images that lack color information. Most existing methods of colorization require laborious user interaction for scribbles or image segmentation. To eliminate the need for human labor, we develop an automatic image colorization method using epitome. Built upon a generative graphical model, epitome is a condensed image appearance and shape model which also proves to be an effective summary of color information for the colorization task. We train the epitome from the reference images and perform inference in the epitome to colorize grayscale images, rendering better colorization results than previous method in our experiments.
1210.4482
Separation of Reliability and Secrecy in Rate-Limited Secret-Key Generation
cs.IT math.IT
For a discrete or a continuous source model, we study the problem of secret-key generation with one round of rate-limited public communication between two legitimate users. Although we do not provide new bounds on the wiretap secret-key (WSK) capacity for the discrete source model, we use an alternative achievability scheme that may be useful for practical applications. As a side result, we conveniently extend known bounds to the case of a continuous source model. Specifically, we consider a sequential key-generation strategy, that implements a rate-limited reconciliation step to handle reliability, followed by a privacy amplification step performed with extractors to handle secrecy. We prove that such a sequential strategy achieves the best known bounds for the rate-limited WSK capacity (under the assumption of degraded sources in the case of two-way communication). However, we show that, unlike the case of rate-unlimited public communication, achieving the reconciliation capacity in a sequential strategy does not necessarily lead to achieving the best known bounds for the WSK capacity. Consequently, reliability and secrecy can be treated successively but not independently, thereby exhibiting a limitation of sequential strategies for rate-limited public communication. Nevertheless, we provide scenarios for which reliability and secrecy can be treated successively and independently, such as the two-way rate-limited SK capacity, the one-way rate-limited WSK capacity for degraded binary symmetric sources, and the one-way rate-limited WSK capacity for Gaussian degraded sources.
1210.4502
Comparing several heuristics for a packing problem
cs.NE
Packing problems are in general NP-hard, even for simple cases. Since now there are no highly efficient algorithms available for solving packing problems. The two-dimensional bin packing problem is about packing all given rectangular items, into a minimum size rectangular bin, without overlapping. The restriction is that the items cannot be rotated. The current paper is comparing a greedy algorithm with a hybrid genetic algorithm in order to see which technique is better for the given problem. The algorithms are tested on different sizes data.
1210.4505
Coherent Fading Channels Driven by Arbitrary Inputs: Asymptotic Characterization of the Constrained Capacity and Related Information- and Estimation-Theoretic Quantities
cs.IT math.IT
We consider the characterization of the asymptotic behavior of the average minimum mean-squared error (MMSE) and the average mutual information in scalar and vector fading coherent channels, where the receiver knows the exact fading channel state but the transmitter knows only the fading channel distribution, driven by a range of inputs. We construct low-snr and -- at the heart of the novelty of the contribution -- high-snr asymptotic expansions for the average MMSE and the average mutual information for coherent channels subject to Rayleigh fading, Ricean fading or Nakagami fading and driven by discrete inputs (with finite support) or various continuous inputs. We reveal the role that the so-called canonical MMSE in a standard additive white Gaussian noise (AWGN) channel plays in the characterization of the asymptotic behavior of the average MMSE and the average mutual information in a fading coherent channel. We also reveal connections to and generalizations of the MMSE dimension. The most relevant element that enables the construction of these non-trivial expansions is the realization that the integral representation of the estimation- and information- theoretic quantities can be seen as an h-transform of a kernel with a monotonic argument: this enables the use of a novel asymptotic expansion of integrals technique -- the Mellin transform method -- that leads immediately to not only the high-snr but also the low-snr expansions of the average MMSE and -- via the I-MMSE relationship -- to expansions of the average mutual information. We conclude with applications of the results to the characterization and optimization of the constrained capacity of a bank of parallel independent coherent fading channels driven by arbitrary discrete inputs.
1210.4507
Submodularity and Optimality of Fusion Rules in Balanced Binary Relay Trees
cs.IT cs.MA math.IT
We study the distributed detection problem in a balanced binary relay tree, where the leaves of the tree are sensors generating binary messages. The root of the tree is a fusion center that makes the overall decision. Every other node in the tree is a fusion node that fuses two binary messages from its child nodes into a new binary message and sends it to the parent node at the next level. We assume that the fusion nodes at the same level use the same fusion rule. We call a string of fusion rules used at different levels a fusion strategy. We consider the problem of finding a fusion strategy that maximizes the reduction in the total error probability between the sensors and the fusion center. We formulate this problem as a deterministic dynamic program and express the solution in terms of Bellman's equations. We introduce the notion of stringsubmodularity and show that the reduction in the total error probability is a stringsubmodular function. Consequentially, we show that the greedy strategy, which only maximizes the level-wise reduction in the total error probability, is within a factor of the optimal strategy in terms of reduction in the total error probability.
1210.4517
Gaming the Game: Honeypot Venues Against Cheaters in Location-based Social Networks
cs.SI cs.CR
The proliferation of location-based social networks (LBSNs) has provided the community with an abundant source of information that can be exploited and used in many different ways. LBSNs offer a number of conveniences to its participants, such as - but not limited to - a list of places in the vicinity of a user, recommendations for an area never explored before provided by other peers, tracking of friends, monetary rewards in the form of special deals from the venues visited as well as a cheap way of advertisement for the latter. However, service convenience and security have followed disjoint paths in LBSNs and users can misuse the offered features. The major threat for the service providers is that of fake check-ins. Users can easily manipulate the localization module of the underlying application and declare their presence in a counterfeit location. The incentives for these behaviors can be both earning monetary as well as virtual rewards. Therefore, while fake check-ins driven from the former motive can cause monetary losses, those aiming in virtual rewards are also harmful. In particular, they can significantly degrade the services offered from the LBSN providers (such as recommendations) or third parties using these data (e.g., urban planners). In this paper, we propose and analyze a honeypot venue-based solution, enhanced with a challenge-response scheme, that flags users who are generating fake spatial information. We believe that our work will stimulate further research on this important topic and will provide new directions with regards to possible solutions.
1210.4567
Gender identity and lexical variation in social media
cs.CL
We present a study of the relationship between gender, linguistic style, and social networks, using a novel corpus of 14,000 Twitter users. Prior quantitative work on gender often treats this social variable as a female/male binary; we argue for a more nuanced approach. By clustering Twitter users, we find a natural decomposition of the dataset into various styles and topical interests. Many clusters have strong gender orientations, but their use of linguistic resources sometimes directly conflicts with the population-level language statistics. We view these clusters as a more accurate reflection of the multifaceted nature of gendered language styles. Previous corpus-based work has also had little to say about individuals whose linguistic styles defy population-level gender patterns. To identify such individuals, we train a statistical classifier, and measure the classifier confidence for each individual in the dataset. Examining individuals whose language does not match the classifier's model for their gender, we find that they have social networks that include significantly fewer same-gender social connections and that, in general, social network homophily is correlated with the use of same-gender language markers. Pairing computational methods and social theory thus offers a new perspective on how gender emerges as individuals position themselves relative to audiences, topics, and mainstream gender norms.
1210.4596
Optimal Achievable Rates for Interference Networks with Random Codes
cs.IT math.IT
The optimal rate region for interference networks is characterized when encoding is restricted to random code ensembles with superposition coding and time sharing. A simple simultaneous nonunique decoding rule, under which each receiver decodes for the intended message as well as the interfering messages, is shown to achieve this optimal rate region regardless of the relative strengths of signal, interference, and noise. This result implies that the Han-Kobayashi bound, the best known inner bound on the capacity region of the two-user-pair interference channel, cannot be improved merely by using the optimal maximum likelihood decoder.
1210.4601
A Direct Approach to Multi-class Boosting and Extensions
cs.LG
Boosting methods combine a set of moderately accurate weaklearners to form a highly accurate predictor. Despite the practical importance of multi-class boosting, it has received far less attention than its binary counterpart. In this work, we propose a fully-corrective multi-class boosting formulation which directly solves the multi-class problem without dividing it into multiple binary classification problems. In contrast, most previous multi-class boosting algorithms decompose a multi-boost problem into multiple binary boosting problems. By explicitly deriving the Lagrange dual of the primal optimization problem, we are able to construct a column generation-based fully-corrective approach to boosting which directly optimizes multi-class classification performance. The new approach not only updates all weak learners' coefficients at every iteration, but does so in a manner flexible enough to accommodate various loss functions and regularizations. For example, it enables us to introduce structural sparsity through mixed-norm regularization to promote group sparsity and feature sharing. Boosting with shared features is particularly beneficial in complex prediction problems where features can be expensive to compute. Our experiments on various data sets demonstrate that our direct multi-class boosting generalizes as well as, or better than, a range of competing multi-class boosting methods. The end result is a highly effective and compact ensemble classifier which can be trained in a distributed fashion.
1210.4614
Random Sequences Based on the Divisor Pairs Function
cs.CR cs.IT math.IT
This paper investigates the randomness properties of a function of the divisor pairs of a natural number. This function, the antecedents of which go to very ancient times, has randomness properties that can find applications in cryptography, key distribution, and other problems of computer science. It is shown that the function is aperiodic and it has excellent autocorrelation properties.
1210.4643
Econoinformatics meets Data-Centric Social Sciences
q-fin.GN cs.SI physics.soc-ph
Our society has been computerised and globalised due to emergence and spread of information and communication technology (ICT). This enables us to investigate our own socio-economic systems based on large amounts of data on human activities. In this article, methods of treating complexity arising from a vast amount of data, and linking data from different sources, are discussed. Furthermore, several examples are given of studies into the applications of econoinformatics for the Japanese stock exchange, foreign exchange markets, domestic hotel booking data and international flight booking data are shown. It is the main message that spatio-temporal information is a key element to synthesise data from different data sources.
1210.4657
Mean-Field Learning: a Survey
cs.LG cs.GT cs.MA math.DS stat.ML
In this paper we study iterative procedures for stationary equilibria in games with large number of players. Most of learning algorithms for games with continuous action spaces are limited to strict contraction best reply maps in which the Banach-Picard iteration converges with geometrical convergence rate. When the best reply map is not a contraction, Ishikawa-based learning is proposed. The algorithm is shown to behave well for Lipschitz continuous and pseudo-contractive maps. However, the convergence rate is still unsatisfactory. Several acceleration techniques are presented. We explain how cognitive users can improve the convergence rate based only on few number of measurements. The methodology provides nice properties in mean field games where the payoff function depends only on own-action and the mean of the mean-field (first moment mean-field games). A learning framework that exploits the structure of such games, called, mean-field learning, is proposed. The proposed mean-field learning framework is suitable not only for games but also for non-convex global optimization problems. Then, we introduce mean-field learning without feedback and examine the convergence to equilibria in beauty contest games, which have interesting applications in financial markets. Finally, we provide a fully distributed mean-field learning and its speedup versions for satisfactory solution in wireless networks. We illustrate the convergence rate improvement with numerical examples.
1210.4663
A PRQ Search Method for Probabilistic Objects
cs.DB cs.CG cs.DS
This article proposes an PQR search method for probabilistic objects. The main idea of our method is to use a strategy called \textit{pre-approximation} that can reduce the initial problem to a highly simplified version, implying that it makes the rest of steps easy to tackle. In particular, this strategy itself is pretty simple and easy to implement. Furthermore, motivated by the cost analysis, we further optimize our solution. The optimizations are mainly based on two insights: (\romannumeral 1) the number of \textit{effective subdivision}s is no more than 1; and (\romannumeral 2) an entity with the larger \textit{span} is more likely to subdivide a single region. We demonstrate the effectiveness and efficiency of our proposed approaches through extensive experiments under various experimental settings.
1210.4695
Regulating the information in spikes: a useful bias
q-bio.NC cs.IT cs.LG math.IT
The bias/variance tradeoff is fundamental to learning: increasing a model's complexity can improve its fit on training data, but potentially worsens performance on future samples. Remarkably, however, the human brain effortlessly handles a wide-range of complex pattern recognition tasks. On the basis of these conflicting observations, it has been argued that useful biases in the form of "generic mechanisms for representation" must be hardwired into cortex (Geman et al). This note describes a useful bias that encourages cooperative learning which is both biologically plausible and rigorously justified.
1210.4700
Optimal Lempel-Ziv based lossy compression for memoryless data: how to make the right mistakes
cs.IT math.IT
Compression refers to encoding data using bits, so that the representation uses as few bits as possible. Compression could be lossless: i.e. encoded data can be recovered exactly from its representation) or lossy where the data is compressed more than the lossless case, but can still be recovered to within prespecified distortion metric. In this paper, we prove the optimality of Codelet Parsing, a quasi-linear time algorithm for lossy compression of sequences of bits that are independently and identically distributed (\iid) and Hamming distortion. Codelet Parsing extends the lossless Lempel Ziv algorithm to the lossy case---a task that has been a focus of the source coding literature for better part of two decades now. Given \iid sequences $\x$, the expected length of the shortest lossy representation such that $\x$ can be reconstructed to within distortion $\dist$ is given by the rate distortion function, $\rd$. We prove the optimality of the Codelet Parsing algorithm for lossy compression of memoryless bit sequences. It splits the input sequence naturally into phrases, representing each phrase by a codelet, a potentially distorted phrase of the same length. The codelets in the lossy representation of a length-$n$ string ${\x}$ have length roughly $(\log n)/\rd$, and like the lossless Lempel Ziv algorithm, Codelet Parsing constructs codebooks logarithmic in the sequence length.
1210.4749
An Auction Approach to Distributed Power Allocation for Multiuser Cooperative Networks
cs.NI cs.GT cs.IT math.IT
This paper studies a wireless network where multiple users cooperate with each other to improve the overall network performance. Our goal is to design an optimal distributed power allocation algorithm that enables user cooperation, in particular, to guide each user on the decision of transmission mode selection and relay selection. Our algorithm has the nice interpretation of an auction mechanism with multiple auctioneers and multiple bidders. Specifically, in our proposed framework, each user acts as both an auctioneer (seller) and a bidder (buyer). Each auctioneer determines its trading price and allocates power to bidders, and each bidder chooses the demand from each auctioneer. By following the proposed distributed algorithm, each user determines how much power to reserve for its own transmission, how much power to purchase from other users, and how much power to contribute for relaying the signals of others. We derive the optimal bidding and pricing strategies that maximize the weighted sum rates of the users. Extensive simulations are carried out to verify our proposed approach.
1210.4752
Discrete Signal Processing on Graphs
cs.SI physics.soc-ph
In social settings, individuals interact through webs of relationships. Each individual is a node in a complex network (or graph) of interdependencies and generates data, lots of data. We label the data by its source, or formally stated, we index the data by the nodes of the graph. The resulting signals (data indexed by the nodes) are far removed from time or image signals indexed by well ordered time samples or pixels. DSP, discrete signal processing, provides a comprehensive, elegant, and efficient methodology to describe, represent, transform, analyze, process, or synthesize these well ordered time or image signals. This paper extends to signals on graphs DSP and its basic tenets, including filters, convolution, z-transform, impulse response, spectral representation, Fourier transform, frequency response, and illustrates DSP on graphs by classifying blogs, linear predicting and compressing data from irregularly located weather stations, or predicting behavior of customers of a mobile service provider.
1210.4759
From Unbalanced Initial Occupant Distribution to Balanced Exit Usage in a Simulation Model of Pedestrian Dynamics
physics.soc-ph cs.MA
It is tested in this contribution if and to which extend a method of a pedestrian simulation tool that attempts to make pedestrians walk into the direction of estimated earliest arrival can help to automatically distribute pedestrians - who are initially distributed arbitrarily in the scenario - equally on the various exits of the scenario.
1210.4778
Average Consensus in the Presence of Delays and Dynamically Changing Directed Graph Topologies
cs.MA cs.DC
Classical approaches for asymptotic convergence to the global average in a distributed fashion typically assume timely and reliable exchange of information between neighboring components of a given multi-component system. These assumptions are not necessarily valid in practical settings due to varying delays that might affect transmissions at different times, as well as possible changes in the underlying interconnection topology (e.g., due to component mobility). In this work, we propose protocols to overcome these limitations. We first consider a fixed interconnection topology (captured by a - possibly directed - graph) and propose a discrete-time protocol that can reach asymptotic average consensus in a distributed fashion, despite the presence of arbitrary (but bounded) delays in the communication links. The protocol requires that each component has knowledge of the number of its outgoing links (i.e., the number of components to which it sends information). We subsequently extend the protocol to also handle changes in the underlying interconnection topology and describe a variety of rather loose conditions under which the modified protocol allows the components to reach asymptotic average consensus. The proposed algorithms are illustrated via examples.
1210.4787
Approximating Acceptance Probabilities of CTMC-Paths on Multi-Clock Deterministic Timed Automata
cs.SY cs.FL
We consider the problem of approximating the probability mass of the set of timed paths under a continuous-time Markov chain (CTMC) that are accepted by a deterministic timed automaton (DTA). As opposed to several existing works on this topic, we consider DTA with multiple clocks. Our key contribution is an algorithm to approximate these probabilities using finite difference methods. An error bound is provided which indicates the approximation error. The stepping stones towards this result include rigorous proofs for the measurability of the set of accepted paths and the integral-equation system characterizing the acceptance probability, and a differential characterization for the acceptance probability.
1210.4791
A computational formulation for constrained solid and liquid membranes considering isogeometric finite elements
cs.CE physics.comp-ph
A geometrically exact membrane formulation is presented that is based on curvilinear coordinates and isogeometric finite elements, and is suitable for both solid and liquid membranes. The curvilinear coordinate system is used to describe both the theory and the finite element equations of the membrane. In the latter case this avoids the use of local cartesian coordinates at the element level. Consequently, no transformation of derivatives is required. The formulation considers a split of the in-plane and out-of-plane membrane contributions, which allows the construction of a stable formulation for liquid membranes with constant surface tension. The proposed membrane formulation is general, and accounts for dead and live loading, as well as enclosed volume, area, and contact constraints. The new formulation is illustrated by several challenging examples, considering linear and quadratic Lagrange elements, as well as isogeometric elements based on quadratic NURBS and cubic T-splines. It is seen that the isogeometric elements are much more accurate than standard Lagrange elements. The gain is especially large for the liquid membrane formulation since it depends explicitly on the surface curvature.
1210.4792
Scalable Matrix-valued Kernel Learning for High-dimensional Nonlinear Multivariate Regression and Granger Causality
stat.ML cs.LG
We propose a general matrix-valued multiple kernel learning framework for high-dimensional nonlinear multivariate regression problems. This framework allows a broad class of mixed norm regularizers, including those that induce sparsity, to be imposed on a dictionary of vector-valued Reproducing Kernel Hilbert Spaces. We develop a highly scalable and eigendecomposition-free algorithm that orchestrates two inexact solvers for simultaneously learning both the input and output components of separable matrix-valued kernels. As a key application enabled by our framework, we show how high-dimensional causal inference tasks can be naturally cast as sparse function estimation problems, leading to novel nonlinear extensions of a class of Graphical Granger Causality techniques. Our algorithmic developments and extensive empirical studies are complemented by theoretical analyses in terms of Rademacher generalization bounds.
1210.4795
Full Rank Solutions for the MIMO Gaussian Wiretap Channel with an Average Power Constraint
cs.IT math.IT
This paper considers a multiple-input multiple-output (MIMO) Gaussian wiretap channel model, where there exists a transmitter, a legitimate receiver and an eavesdropper, each equipped with multiple antennas. In this paper, we first revisit the rank property of the optimal input covariance matrix that achieves the secrecy capacity of the multiple antenna MIMO Gaussian wiretap channel under the average power constraint. Next, we obtain necessary and sufficient conditions on the MIMO wiretap channel parameters such that the optimal input covariance matrix is full-rank, and we fully characterize the resulting covariance matrix as well. Numerical results are presented to illustrate the proposed theoretical findings.
1210.4808
Basic Experiment Planning via Information Metrics: the RoboMendel Problem
cs.IT math.IT
In this paper we outline some mathematical questions that emerge from trying to "turn the scientific method into math". Specifically, we consider the problem of experiment planning (choosing the best experiment to do next) in explicit probabilistic and information theoretic terms. We formulate this as an information measurement problem; that is, we seek a rigorous definition of an information metric to measure the likely information yield of an experiment, such that maximizing the information metric will indeed reliably choose the best experiment to perform. We present the surprising result that defining the metric purely in terms of prediction power on observable variables yields a metric that can converge to the classical mutual information measuring how informative the experimental observation is about an underlying hidden variable. We show how the expectation potential information metric can compute the "information rate" of an experiment as well its total possible yield, and the information value of experimental controls. To illustrate the utility of these concepts for guiding fundamental scientific inquiry, we present an extensive case study (RoboMendel) applying these metrics to propose sequences of experiments for discovering the basic principles of genetics.
1210.4831
Closing the Gap to the Capacity of APSK: Constellation Shaping and Degree Distributions
cs.IT math.IT
Constellation shaping is an energy-efficient strategy involving the transmission of lower-energy signals more frequently than higher-energy signals. Previous work has shown that shaping is particularly effective when used with coded amplitude phase-shift keying (APSK), a modulation that has been popularized recently due to its inclusion in the DVB-S2 standard. While shaped APSK can provide significant gains when used with standard off-the-shelf LDPC codes, such as the codes in the DVB-S2 standard, additional non-negligible gains can be achieved by optimizing the LDPC code with respect to the shaped APSK modulation. In this paper, we optimize the degree distributions of the LDPC code used in conjunction with shaped APSK. The optimization process is an extension of the EXIT-chart technique of ten Brink, et al., which has been adapted to account for the shaped APSK modulation. We begin by constraining the code to have the same number of distinct variable-node degrees as the codes in the DVB-S2 standard, and show that the optimization provides 32-APSK systems with an additional coding gain of 0.34 dB at a system rate of R=3 bits per symbol, compared to shaped systems that use the long LDPC code from the DVB-S2 standard. We then increase the number of allowed variable node degrees by one, and find that an additional 0.1 dB gain is achievable.
1210.4839
Leveraging Side Observations in Stochastic Bandits
cs.LG stat.ML
This paper considers stochastic bandits with side observations, a model that accounts for both the exploration/exploitation dilemma and relationships between arms. In this setting, after pulling an arm i, the decision maker also observes the rewards for some other actions related to i. We will see that this model is suited to content recommendation in social networks, where users' reactions may be endorsed or not by their friends. We provide efficient algorithms based on upper confidence bounds (UCBs) to leverage this additional information and derive new bounds improving on standard regret guarantees. We also evaluate these policies in the context of movie recommendation in social networks: experiments on real datasets show substantial learning rate speedups ranging from 2.2x to 14x on dense networks.
1210.4840
Lifted Relax, Compensate and then Recover: From Approximate to Exact Lifted Probabilistic Inference
cs.AI
We propose an approach to lifted approximate inference for first-order probabilistic models, such as Markov logic networks. It is based on performing exact lifted inference in a simplified first-order model, which is found by relaxing first-order constraints, and then compensating for the relaxation. These simplified models can be incrementally improved by carefully recovering constraints that have been relaxed, also at the first-order level. This leads to a spectrum of approximations, with lifted belief propagation on one end, and exact lifted inference on the other. We discuss how relaxation, compensation, and recovery can be performed, all at the firstorder level, and show empirically that our approach substantially improves on the approximations of both propositional solvers and lifted belief propagation.
1210.4841
An Efficient Message-Passing Algorithm for the M-Best MAP Problem
cs.AI cs.LG stat.ML
Much effort has been directed at algorithms for obtaining the highest probability configuration in a probabilistic random field model known as the maximum a posteriori (MAP) inference problem. In many situations, one could benefit from having not just a single solution, but the top M most probable solutions known as the M-Best MAP problem. In this paper, we propose an efficient message-passing based algorithm for solving the M-Best MAP problem. Specifically, our algorithm solves the recently proposed Linear Programming (LP) formulation of M-Best MAP [7], while being orders of magnitude faster than a generic LP-solver. Our approach relies on studying a particular partial Lagrangian relaxation of the M-Best MAP LP which exposes a natural combinatorial structure of the problem that we exploit.
1210.4842
Causal Inference by Surrogate Experiments: z-Identifiability
cs.AI stat.ME
We address the problem of estimating the effect of intervening on a set of variables X from experiments on a different set, Z, that is more accessible to manipulation. This problem, which we call z-identifiability, reduces to ordinary identifiability when Z = empty and, like the latter, can be given syntactic characterization using the do-calculus [Pearl, 1995; 2000]. We provide a graphical necessary and sufficient condition for z-identifiability for arbitrary sets X,Z, and Y (the outcomes). We further develop a complete algorithm for computing the causal effect of X on Y using information provided by experiments on Z. Finally, we use our results to prove completeness of do-calculus relative to z-identifiability, a result that does not follow from completeness relative to ordinary identifiability.
1210.4843
Deterministic MDPs with Adversarial Rewards and Bandit Feedback
cs.GT cs.LG
We consider a Markov decision process with deterministic state transition dynamics, adversarially generated rewards that change arbitrarily from round to round, and a bandit feedback model in which the decision maker only observes the rewards it receives. In this setting, we present a novel and efficient online decision making algorithm named MarcoPolo. Under mild assumptions on the structure of the transition dynamics, we prove that MarcoPolo enjoys a regret of O(T^(3/4)sqrt(log(T))) against the best deterministic policy in hindsight. Specifically, our analysis does not rely on the stringent unichain assumption, which dominates much of the previous work on this topic.
1210.4845
Exploiting Uniform Assignments in First-Order MPE
cs.AI
The MPE (Most Probable Explanation) query plays an important role in probabilistic inference. MPE solution algorithms for probabilistic relational models essentially adapt existing belief assessment method, replacing summation with maximization. But the rich structure and symmetries captured by relational models together with the properties of the maximization operator offer an opportunity for additional simplification with potentially significant computational ramifications. Specifically, these models often have groups of variables that define symmetric distributions over some population of formulas. The maximizing choice for different elements of this group is the same. If we can realize this ahead of time, we can significantly reduce the size of the model by eliminating a potentially significant portion of random variables. This paper defines the notion of uniformly assigned and partially uniformly assigned sets of variables, shows how one can recognize these sets efficiently, and how the model can be greatly simplified once we recognize them, with little computational effort. We demonstrate the effectiveness of these ideas empirically on a number of models.
1210.4846
Variational Dual-Tree Framework for Large-Scale Transition Matrix Approximation
cs.LG stat.ML
In recent years, non-parametric methods utilizing random walks on graphs have been used to solve a wide range of machine learning problems, but in their simplest form they do not scale well due to the quadratic complexity. In this paper, a new dual-tree based variational approach for approximating the transition matrix and efficiently performing the random walk is proposed. The approach exploits a connection between kernel density estimation, mixture modeling, and random walk on graphs in an optimization of the transition matrix for the data graph that ties together edge transitions probabilities that are similar. Compared to the de facto standard approximation method based on k-nearestneighbors, we demonstrate order of magnitudes speedup without sacrificing accuracy for Label Propagation tasks on benchmark data sets in semi-supervised learning.
1210.4848
Uncertain Congestion Games with Assorted Human Agent Populations
cs.GT cs.MA
Congestion games model a wide variety of real-world resource congestion problems, such as selfish network routing, traffic route guidance in congested areas, taxi fleet optimization and crowd movement in busy areas. However, existing research in congestion games assumes: (a) deterministic movement of agents between resources; and (b) perfect rationality (i.e. maximizing their own expected value) of all agents. Such assumptions are not reasonable in dynamic domains where decision support has to be provided to humans. For instance, in optimizing the performance of a taxi fleet serving a city, movement of taxis can be involuntary or nondeterministic (decided by the specific customer who hires the taxi) and more importantly, taxi drivers may not follow advice provided by the decision support system (due to bounded rationality of humans). To that end, we contribute: (a) a general framework for representing congestion games under uncertainty for populations with assorted notions of rationality. (b) a scalable approach for solving the decision problem for perfectly rational agents which are in the mix with boundedly rational agents; and (c) a detailed evaluation on a synthetic and realworld data set to illustrate the usefulness of our new approach with respect to key social welfare metrics in the context of an assorted human-agent population. An interesting result from our experiments on a real-world taxi fleet optimization problem is that it is better (in terms of revenue and operational efficiency) for taxi drivers to follow perfectly rational strategies irrespective of the percentage of drivers not following the advice.
1210.4849
Toward Large-Scale Agent Guidance in an Urban Taxi Service
cs.MA cs.AI cs.GT
Empty taxi cruising represents a wastage of resources in the context of urban taxi services. In this work, we seek to minimize such wastage. An analysis of a large trace of taxi operations reveals that the services' inefficiency is caused by drivers' greedy cruising behavior. We model the existing system as a continuous time Markov chain. To address the problem, we propose that each taxi be equipped with an intelligent agent that will guide the driver when cruising for passengers. Then, drawing from AI literature on multiagent planning, we explore two possible ways to compute such guidance. The first formulation assumes fully cooperative drivers. This allows us, in principle, to compute systemwide optimal cruising policy. This is modeled as a Markov decision process. The second formulation assumes rational drivers, seeking to maximize their own profit. This is modeled as a stochastic congestion game, a specialization of stochastic games. Nash equilibrium policy is proposed as the solution to the game, where no driver has the incentive to singly deviate from it. Empirical result shows that both formulations improve the efficiency of the service significantly.
1210.4850
Markov Determinantal Point Processes
cs.LG cs.IR stat.ML
A determinantal point process (DPP) is a random process useful for modeling the combinatorial problem of subset selection. In particular, DPPs encourage a random subset Y to contain a diverse set of items selected from a base set Y. For example, we might use a DPP to display a set of news headlines that are relevant to a user's interests while covering a variety of topics. Suppose, however, that we are asked to sequentially select multiple diverse sets of items, for example, displaying new headlines day-by-day. We might want these sets to be diverse not just individually but also through time, offering headlines today that are unlike the ones shown yesterday. In this paper, we construct a Markov DPP (M-DPP) that models a sequence of random sets {Yt}. The proposed M-DPP defines a stationary process that maintains DPP margins. Crucially, the induced union process Zt = Yt u Yt-1 is also marginally DPP-distributed. Jointly, these properties imply that the sequence of random sets are encouraged to be diverse both at a given time step as well as across time steps. We describe an exact, efficient sampling procedure, and a method for incrementally learning a quality measure over items in the base set Y based on external preferences. We apply the M-DPP to the task of sequentially displaying diverse and relevant news articles to a user with topic preferences.
1210.4851
Learning to Rank With Bregman Divergences and Monotone Retargeting
cs.LG stat.ML
This paper introduces a novel approach for learning to rank (LETOR) based on the notion of monotone retargeting. It involves minimizing a divergence between all monotonic increasing transformations of the training scores and a parameterized prediction function. The minimization is both over the transformations as well as over the parameters. It is applied to Bregman divergences, a large class of "distance like" functions that were recently shown to be the unique class that is statistically consistent with the normalized discounted gain (NDCG) criterion [19]. The algorithm uses alternating projection style updates, in which one set of simultaneous projections can be computed independent of the Bregman divergence and the other reduces to parameter estimation of a generalized linear model. This results in easily implemented, efficiently parallelizable algorithm for the LETOR task that enjoys global optimum guarantees under mild conditions. We present empirical results on benchmark datasets showing that this approach can outperform the state of the art NDCG consistent techniques.
1210.4852
The Do-Calculus Revisited
cs.AI stat.ME
The do-calculus was developed in 1995 to facilitate the identification of causal effects in non-parametric models. The completeness proofs of [Huang and Valtorta, 2006] and [Shpitser and Pearl, 2006] and the graphical criteria of [Tian and Shpitser, 2010] have laid this identification problem to rest. Recent explorations unveil the usefulness of the do-calculus in three additional areas: mediation analysis [Pearl, 2012], transportability [Pearl and Bareinboim, 2011] and metasynthesis. Meta-synthesis (freshly coined) is the task of fusing empirical results from several diverse studies, conducted on heterogeneous populations and under different conditions, so as to synthesize an estimate of a causal relation in some target environment, potentially different from those under study. The talk surveys these results with emphasis on the challenges posed by meta-synthesis. For background material, see http://bayes.cs.ucla.edu/csl_papers.html
1210.4853
Weighted Sets of Probabilities and MinimaxWeighted Expected Regret: New Approaches for Representing Uncertainty and Making Decisions
cs.GT cs.AI q-fin.TR
We consider a setting where an agent's uncertainty is represented by a set of probability measures, rather than a single measure. Measure-bymeasure updating of such a set of measures upon acquiring new information is well-known to suffer from problems; agents are not always able to learn appropriately. To deal with these problems, we propose using weighted sets of probabilities: a representation where each measure is associated with a weight, which denotes its significance. We describe a natural approach to updating in such a situation and a natural approach to determining the weights. We then show how this representation can be used in decision-making, by modifying a standard approach to decision making-minimizing expected regret-to obtain minimax weighted expected regret (MWER).We provide an axiomatization that characterizes preferences induced by MWER both in the static and dynamic case.
1210.4854
Semantic Understanding of Professional Soccer Commentaries
cs.CL cs.AI
This paper presents a novel approach to the problem of semantic parsing via learning the correspondences between complex sentences and rich sets of events. Our main intuition is that correct correspondences tend to occur more frequently. Our model benefits from a discriminative notion of similarity to learn the correspondence between sentence and an event and a ranking machinery that scores the popularity of each correspondence. Our method can discover a group of events (called macro-events) that best describes a sentence. We evaluate our method on our novel dataset of professional soccer commentaries. The empirical results show that our method significantly outperforms the state-of-theart.
1210.4855
A Slice Sampler for Restricted Hierarchical Beta Process with Applications to Shared Subspace Learning
cs.LG cs.CV stat.ML
Hierarchical beta process has found interesting applications in recent years. In this paper we present a modified hierarchical beta process prior with applications to hierarchical modeling of multiple data sources. The novel use of the prior over a hierarchical factor model allows factors to be shared across different sources. We derive a slice sampler for this model, enabling tractable inference even when the likelihood and the prior over parameters are non-conjugate. This allows the application of the model in much wider contexts without restrictions. We present two different data generative models a linear GaussianGaussian model for real valued data and a linear Poisson-gamma model for count data. Encouraging transfer learning results are shown for two real world applications text modeling and content based image retrieval.
1210.4856
Exploiting compositionality to explore a large space of model structures
cs.LG stat.ML
The recent proliferation of richly structured probabilistic models raises the question of how to automatically determine an appropriate model for a dataset. We investigate this question for a space of matrix decomposition models which can express a variety of widely used models from unsupervised learning. To enable model selection, we organize these models into a context-free grammar which generates a wide variety of structures through the compositional application of a few simple rules. We use our grammar to generically and efficiently infer latent components and estimate predictive likelihood for nearly 2500 structures using a small toolbox of reusable algorithms. Using a greedy search over our grammar, we automatically choose the decomposition structure from raw data by evaluating only a small fraction of all models. The proposed method typically finds the correct structure for synthetic data and backs off gracefully to simpler models under heavy noise. It learns sensible structures for datasets as diverse as image patches, motion capture, 20 Questions, and U.S. Senate votes, all using exactly the same code.
1210.4857
Generalized Belief Propagation on Tree Robust Structured Region Graphs
cs.AI
This paper provides some new guidance in the construction of region graphs for Generalized Belief Propagation (GBP). We connect the problem of choosing the outer regions of a LoopStructured Region Graph (SRG) to that of finding a fundamental cycle basis of the corresponding Markov network. We also define a new class of tree-robust Loop-SRG for which GBP on any induced (spanning) tree of the Markov network, obtained by setting to zero the off-tree interactions, is exact. This class of SRG is then mapped to an equivalent class of tree-robust cycle bases on the Markov network. We show that a treerobust cycle basis can be identified by proving that for every subset of cycles, the graph obtained from the edges that participate in a single cycle only, is multiply connected. Using this we identify two classes of tree-robust cycle bases: planar cycle bases and "star" cycle bases. In experiments we show that tree-robustness can be successfully exploited as a design principle to improve the accuracy and convergence of GBP.
1210.4859
Mechanism Design for Cost Optimal PAC Learning in the Presence of Strategic Noisy Annotators
cs.LG cs.GT stat.ML
We consider the problem of Probably Approximate Correct (PAC) learning of a binary classifier from noisy labeled examples acquired from multiple annotators (each characterized by a respective classification noise rate). First, we consider the complete information scenario, where the learner knows the noise rates of all the annotators. For this scenario, we derive sample complexity bound for the Minimum Disagreement Algorithm (MDA) on the number of labeled examples to be obtained from each annotator. Next, we consider the incomplete information scenario, where each annotator is strategic and holds the respective noise rate as a private information. For this scenario, we design a cost optimal procurement auction mechanism along the lines of Myerson's optimal auction design framework in a non-trivial manner. This mechanism satisfies incentive compatibility property, thereby facilitating the learner to elicit true noise rates of all the annotators.
1210.4860
Spectral Estimation of Conditional Random Graph Models for Large-Scale Network Data
cs.SI cs.LG physics.soc-ph stat.ML
Generative models for graphs have been typically committed to strong prior assumptions concerning the form of the modeled distributions. Moreover, the vast majority of currently available models are either only suitable for characterizing some particular network properties (such as degree distribution or clustering coefficient), or they are aimed at estimating joint probability distributions, which is often intractable in large-scale networks. In this paper, we first propose a novel network statistic, based on the Laplacian spectrum of graphs, which allows to dispense with any parametric assumption concerning the modeled network properties. Second, we use the defined statistic to develop the Fiedler random graph model, switching the focus from the estimation of joint probability distributions to a more tractable conditional estimation setting. After analyzing the dependence structure characterizing Fiedler random graphs, we evaluate them experimentally in edge prediction over several real-world networks, showing that they allow to reach a much higher prediction accuracy than various alternative statistical models.
1210.4861
Uniform Solution Sampling Using a Constraint Solver As an Oracle
cs.AI
We consider the problem of sampling from solutions defined by a set of hard constraints on a combinatorial space. We propose a new sampling technique that, while enforcing a uniform exploration of the search space, leverages the reasoning power of a systematic constraint solver in a black-box scheme. We present a series of challenging domains, such as energy barriers and highly asymmetric spaces, that reveal the difficulties introduced by hard constraints. We demonstrate that standard approaches such as Simulated Annealing and Gibbs Sampling are greatly affected, while our new technique can overcome many of these difficulties. Finally, we show that our sampling scheme naturally defines a new approximate model counting technique, which we empirically show to be very accurate on a range of benchmark problems.
1210.4862
Sample-efficient Nonstationary Policy Evaluation for Contextual Bandits
cs.LG stat.ML
We present and prove properties of a new offline policy evaluator for an exploration learning setting which is superior to previous evaluators. In particular, it simultaneously and correctly incorporates techniques from importance weighting, doubly robust evaluation, and nonstationary policy evaluation approaches. In addition, our approach allows generating longer histories by careful control of a bias-variance tradeoff, and further decreases variance by incorporating information about randomness of the target policy. Empirical evidence from synthetic and realworld exploration learning problems shows the new evaluator successfully unifies previous approaches and uses information an order of magnitude more efficiently.
1210.4863
DBN-Based Combinatorial Resampling for Articulated Object Tracking
cs.CV
Particle Filter is an effective solution to track objects in video sequences in complex situations. Its key idea is to estimate the density over the possible states of the object using a weighted sample whose elements are called particles. One of its crucial step is a resampling step in which particles are resampled to avoid some degeneracy problem. In this paper, we introduce a new resampling method called Combinatorial Resampling that exploits some features of articulated objects to resample over an implicitly created sample of an exponential size better representing the density to estimate. We prove that it is sound and, through experimentations both on challenging synthetic and real video sequences, we show that it outperforms all classical resampling methods both in terms of the quality of its results and in terms of response times.
1210.4864
Graph-Coupled HMMs for Modeling the Spread of Infection
cs.SI physics.soc-ph stat.AP
We develop Graph-Coupled Hidden Markov Models (GCHMMs) for modeling the spread of infectious disease locally within a social network. Unlike most previous research in epidemiology, which typically models the spread of infection at the level of entire populations, we successfully leverage mobile phone data collected from 84 people over an extended period of time to model the spread of infection on an individual level. Our model, the GCHMM, is an extension of widely-used Coupled Hidden Markov Models (CHMMs), which allow dependencies between state transitions across multiple Hidden Markov Models (HMMs), to situations in which those dependencies are captured through the structure of a graph, or to social networks that may change over time. The benefit of making infection predictions on an individual level is enormous, as it allows people to receive more personalized and relevant health advice.
1210.4865
Scaling Up Decentralized MDPs Through Heuristic Search
cs.AI cs.MA
Decentralized partially observable Markov decision processes (Dec-POMDPs) are rich models for cooperative decision-making under uncertainty, but are often intractable to solve optimally (NEXP-complete). The transition and observation independent Dec-MDP is a general subclass that has been shown to have complexity in NP, but optimal algorithms for this subclass are still inefficient in practice. In this paper, we first provide an updated proof that an optimal policy does not depend on the histories of the agents, but only the local observations. We then present a new algorithm based on heuristic search that is able to expand search nodes by using constraint optimization. We show experimental results comparing our approach with the state-of-the-art DecMDP and Dec-POMDP solvers. These results show a reduction in computation time and an increase in scalability by multiple orders of magnitude in a number of benchmarks.
1210.4866
A Bayesian Approach to Constraint Based Causal Inference
cs.AI stat.ME
We target the problem of accuracy and robustness in causal inference from finite data sets. Some state-of-the-art algorithms produce clear output complete with solid theoretical guarantees but are susceptible to propagating erroneous decisions, while others are very adept at handling and representing uncertainty, but need to rely on undesirable assumptions. Our aim is to combine the inherent robustness of the Bayesian approach with the theoretical strength and clarity of constraint-based methods. We use a Bayesian score to obtain probability estimates on the input statements used in a constraint-based procedure. These are subsequently processed in decreasing order of reliability, letting more reliable decisions take precedence in case of con icts, until a single output model is obtained. Tests show that a basic implementation of the resulting Bayesian Constraint-based Causal Discovery (BCCD) algorithm already outperforms established procedures such as FCI and Conservative PC. It can also indicate which causal decisions in the output have high reliability and which do not.
1210.4867
Lifted Relational Variational Inference
cs.LG stat.ML
Hybrid continuous-discrete models naturally represent many real-world applications in robotics, finance, and environmental engineering. Inference with large-scale models is challenging because relational structures deteriorate rapidly during inference with observations. The main contribution of this paper is an efficient relational variational inference algorithm that factors largescale probability models into simpler variational models, composed of mixtures of iid (Bernoulli) random variables. The algorithm takes probability relational models of largescale hybrid systems and converts them to a close-to-optimal variational models. Then, it efficiently calculates marginal probabilities on the variational models by using a latent (or lifted) variable elimination or a lifted stochastic sampling. This inference is unique because it maintains the relational structure upon individual observations and during inference steps.
1210.4868
Graphical-model Based Multiple Testing under Dependence, with Applications to Genome-wide Association Studies
stat.ME cs.CE stat.AP
Large-scale multiple testing tasks often exhibit dependence, and leveraging the dependence between individual tests is still one challenging and important problem in statistics. With recent advances in graphical models, it is feasible to use them to perform multiple testing under dependence. We propose a multiple testing procedure which is based on a Markov-random-field-coupled mixture model. The ground truth of hypotheses is represented by a latent binary Markov random field, and the observed test statistics appear as the coupled mixture variables. The parameters in our model can be automatically learned by a novel EM algorithm. We use an MCMC algorithm to infer the posterior probability that each hypothesis is null (termed local index of significance), and the false discovery rate can be controlled accordingly. Simulations show that the numerical performance of multiple testing can be improved substantially by using our procedure. We apply the procedure to a real-world genome-wide association study on breast cancer, and we identify several SNPs with strong association evidence.
1210.4869
Response Aware Model-Based Collaborative Filtering
cs.LG cs.IR stat.ML
Previous work on recommender systems mainly focus on fitting the ratings provided by users. However, the response patterns, i.e., some items are rated while others not, are generally ignored. We argue that failing to observe such response patterns can lead to biased parameter estimation and sub-optimal model performance. Although several pieces of work have tried to model users' response patterns, they miss the effectiveness and interpretability of the successful matrix factorization collaborative filtering approaches. To bridge the gap, in this paper, we unify explicit response models and PMF to establish the Response Aware Probabilistic Matrix Factorization (RAPMF) framework. We show that RAPMF subsumes PMF as a special case. Empirically we demonstrate the merits of RAPMF from various aspects.
1210.4870
Crowdsourcing Control: Moving Beyond Multiple Choice
cs.AI cs.LG
To ensure quality results from crowdsourced tasks, requesters often aggregate worker responses and use one of a plethora of strategies to infer the correct answer from the set of noisy responses. However, all current models assume prior knowledge of all possible outcomes of the task. While not an unreasonable assumption for tasks that can be posited as multiple-choice questions (e.g. n-ary classification), we observe that many tasks do not naturally fit this paradigm, but instead demand a free-response formulation where the outcome space is of infinite size (e.g. audio transcription). We model such tasks with a novel probabilistic graphical model, and design and implement LazySusan, a decision-theoretic controller that dynamically requests responses as necessary in order to infer answers to these tasks. We also design an EM algorithm to jointly learn the parameters of our model while inferring the correct answers to multiple tasks at a time. Live experiments on Amazon Mechanical Turk demonstrate the superiority of LazySusan at solving SAT Math questions, eliminating 83.2% of the error and achieving greater net utility compared to the state-ofthe-art strategy, majority-voting. We also show in live experiments that our EM algorithm outperforms majority-voting on a visualization task that we design.
1210.4871
Learning Mixtures of Submodular Shells with Application to Document Summarization
cs.LG cs.CL cs.IR stat.ML
We introduce a method to learn a mixture of submodular "shells" in a large-margin setting. A submodular shell is an abstract submodular function that can be instantiated with a ground set and a set of parameters to produce a submodular function. A mixture of such shells can then also be so instantiated to produce a more complex submodular function. What our algorithm learns are the mixture weights over such shells. We provide a risk bound guarantee when learning in a large-margin structured-prediction setting using a projected subgradient method when only approximate submodular optimization is possible (such as with submodular function maximization). We apply this method to the problem of multi-document summarization and produce the best results reported so far on the widely used NIST DUC-05 through DUC-07 document summarization corpora.
1210.4872
Nested Dictionary Learning for Hierarchical Organization of Imagery and Text
cs.LG cs.CV stat.ML
A tree-based dictionary learning model is developed for joint analysis of imagery and associated text. The dictionary learning may be applied directly to the imagery from patches, or to general feature vectors extracted from patches or superpixels (using any existing method for image feature extraction). Each image is associated with a path through the tree (from root to a leaf), and each of the multiple patches in a given image is associated with one node in that path. Nodes near the tree root are shared between multiple paths, representing image characteristics that are common among different types of images. Moving toward the leaves, nodes become specialized, representing details in image classes. If available, words (text) are also jointly modeled, with a path-dependent probability over words. The tree structure is inferred via a nested Dirichlet process, and a retrospective stick-breaking sampler is used to infer the tree depth and width.
1210.4874
Dynamic Stochastic Orienteering Problems for Risk-Aware Applications
cs.AI cs.DS
Orienteering problems (OPs) are a variant of the well-known prize-collecting traveling salesman problem, where the salesman needs to choose a subset of cities to visit within a given deadline. OPs and their extensions with stochastic travel times (SOPs) have been used to model vehicle routing problems and tourist trip design problems. However, they suffer from two limitations travel times between cities are assumed to be time independent and the route provided is independent of the risk preference (with respect to violating the deadline) of the user. To address these issues, we make the following contributions: We introduce (1) a dynamic SOP (DSOP) model, which is an extension of SOPs with dynamic (time-dependent) travel times; (2) a risk-sensitive criterion to allow for different risk preferences; and (3) a local search algorithm to solve DSOPs with this risk-sensitive criterion. We evaluated our algorithms on a real-world dataset for a theme park navigation problem as well as synthetic datasets employed in the literature.