arXiv subjects

Kush R. Varshney

Publications and source records attributed to Kush R. Varshney.

102 records · Page 6Linked to original sources

arXiv2016

Dynamic matrix factorization with social influence

Matrix factorization is a key component of collaborative filtering-based recommendation systems because it allows us to complete sparse user-by-item ratings matrices under a low-rank assumption that encodes the belief that similar users give similar ratings and that similar items garner similar ratings. This paradigm has had immeasurable practical success, but it is not the complete story for understanding and inferring the preferences of people. First, peoples' preferences and their observable manifestations as ratings evolve over time along general patterns of trajectories. Second, an individual person's preferences evolve over time through influence of their social connections. In this paper, we develop a unified process model for both types of dynamics within a state space approach, together with an efficient optimization scheme for estimation within that model. The model combines elements from recent developments in dynamic matrix factorization, opinion dynamics and social learning, and trust-based recommendation. The estimation builds upon recent advances in numerical nonlinear optimization. Empirical results on a large-scale data set from the Epinions website demonstrate consistent reduction in root mean squared error by consideration of the two types of dynamics.

stat.ML↗

arXiv2016

Engineering Safety in Machine Learning

Machine learning algorithms are increasingly influencing our decisions and interacting with us in all parts of our daily lives. Therefore, just like for power plants, highways, and myriad other engineered sociotechnical systems, we must consider the safety of systems involving machine learning. In this paper, we first discuss the definition of safety in terms of risk, epistemic uncertainty, and the harm incurred by unwanted outcomes. Then we examine dimensions, such as the choice of cost function and the appropriateness of minimizing the empirical average training cost, along which certain real-world applications may not be completely amenable to the foundational principle of modern statistical machine learning: empirical risk minimization. In particular, we note an emerging dichotomy of applications: ones in which safety is important and risk minimization is not the complete story (we name these Type A applications), and ones in which safety is not so critical and risk minimization is sufficient (we name these Type B applications). Finally, we discuss how four different strategies for achieving safety in engineering (inherently safe design, safety reserves, safe fail, and procedural safeguards) can be mapped to the machine learning context through interpretability and causality of predictive models, objectives beyond expected prediction accuracy, human involvement for labeling difficult or rare examples, and user experience design of software.

stat.ML↗

arXiv2015

Interpretable Two-level Boolean Rule Learning for Classification

This paper proposes algorithms for learning two-level Boolean rules in Conjunctive Normal Form (CNF, i.e. AND-of-ORs) or Disjunctive Normal Form (DNF, i.e. OR-of-ANDs) as a type of human-interpretable classification model, aiming for a favorable trade-off between the classification accuracy and the simplicity of the rule. Two formulations are proposed. The first is an integer program whose objective function is a combination of the total number of errors and the total number of features used in the rule. We generalize a previously proposed linear programming (LP) relaxation from one-level to two-level rules. The second formulation replaces the 0-1 classification error with the Hamming distance from the current two-level rule to the closest rule that correctly classifies a sample. Based on this second formulation, block coordinate descent and alternating minimization algorithms are developed. Experiments show that the two-level rules can yield noticeably better performance than one-level rules due to their dramatically larger modeling capacity, and the two algorithms based on the Hamming distance formulation are generally superior to the other two-level rule learning methods in our comparison. A proposed approach to binarize any fractional values in the optimal solutions of LP relaxations is also shown to be effective.

cs.LG↗

arXiv2015

Olfactory Signal Processing

Olfaction, the sense of smell, has received scant attention from a signal processing perspective in comparison to audition and vision. In this paper, we develop a signal processing paradigm for olfactory signals based on new scientific discoveries including the psychophysics concept of olfactory white. We describe a framework for predicting the perception of odorant compounds from their physicochemical features and use the prediction as a foundation for several downstream processing tasks. We detail formulations for odor cancellation and food steganography, and provide real-world empirical examples for the two tasks. We also discuss adaptive filtering and other olfactory signal processing tasks at a high level.

cs.IT↗

arXiv2014

Robust Binary Hypothesis Testing Under Contaminated Likelihoods

In hypothesis testing, the phenomenon of label noise, in which hypothesis labels are switched at random, contaminates the likelihood functions. In this paper, we develop a new method to determine the decision rule when we do not have knowledge of the uncontaminated likelihoods and contamination probabilities, but only have knowledge of the contaminated likelihoods. In particular we pose a minimax optimization problem that finds a decision rule robust against this lack of knowledge. The method simplifies by application of linear programming theory. Motivation for this investigation is provided by problems encountered in workforce analytics.

cs.IT↗

arXiv2013

A Big Data Approach to Computational Creativity

Computational creativity is an emerging branch of artificial intelligence that places computers in the center of the creative process. Broadly, creativity involves a generative step to produce many ideas and a selective step to determine the ones that are the best. Many previous attempts at computational creativity, however, have not been able to achieve a valid selective step. This work shows how bringing data sources from the creative domain and from hedonic psychophysics together with big data analytics techniques can overcome this shortcoming to yield a system that can produce novel and high-quality creative artifacts. Our data-driven approach is demonstrated through a computational creativity system for culinary recipes and menus we developed and deployed, which can operate either autonomously or semi-autonomously with human interaction. We also comment on the volume, velocity, variety, and veracity of data in computational creativity.

cs.CY↗

arXiv2013

Bounded Confidence Opinion Dynamics in a Social Network of Bayesian Decision Makers

Bounded confidence opinion dynamics model the propagation of information in social networks. However in the existing literature, opinions are only viewed as abstract quantities without semantics rather than as part of a decision-making system. In this work, opinion dynamics are examined when agents are Bayesian decision makers that perform hypothesis testing or signal detection, and the dynamics are applied to prior probabilities of hypotheses. Bounded confidence is defined on prior probabilities through Bayes risk error divergence, the appropriate measure between priors in hypothesis testing. This definition contrasts with the measure used between opinions in standard models: absolute error. It is shown that the rapid convergence of prior probabilities to a small number of limiting values is similar to that seen in the standard Krause-Hegselmann model. The most interesting finding in this work is that the number of these limiting values and the time to convergence changes with the signal-to-noise ratio in the detection task. The number of final values or clusters is maximal at intermediate signal-to-noise ratios, suggesting that the most contentious issues lead to the largest number of factions. It is at these same intermediate signal-to-noise ratios at which the degradation in detection performance of the aggregate vote of the decision makers is greatest in comparison to the Bayes optimal detection performance.

cs.SI↗

arXiv2013

Flavor Pairing in Medieval European Cuisine: A Study in Cooking with Dirty Data

An important part of cooking with computers is using statistical methods to create new, flavorful ingredient combinations. The flavor pairing hypothesis states that culinary ingredients with common chemical flavor components combine well to produce pleasant dishes. It has been recently shown that this design principle is a basis for modern Western cuisine and is reversed for Asian cuisine. Such data-driven analysis compares the chemistry of ingredients to ingredient sets found in recipes. However, analytics-based generation of novel flavor profiles can only be as good as the underlying chemical and recipe data. Incomplete, inaccurate, and irrelevant data may degrade flavor pairing inferences. Chemical data on flavor compounds is incomplete due to the nature of the experiments that must be conducted to obtain it. Recipe data may have issues due to text parsing errors, imprecision in textual descriptions of ingredients, and the fact that the same ingredient may be known by different names in different recipes. Moreover, the process of matching ingredients in chemical data and recipe data may be fraught with mistakes. Much of the `dirtiness' of the data cannot be cleansed even with manual curation. In this work, we collect a new data set of recipes from Medieval Europe before the Columbian Exchange and investigate the flavor pairing hypothesis historically. To investigate the role of data incompleteness and error as part of this hypothesis testing, we use two separate chemical compound data sets with different levels of cleanliness. Notably, the different data sets give conflicting conclusions about the flavor pairing hypothesis in Medieval Europe. As a contribution towards social science, we obtain inferences about the evolution of culinary arts when many new ingredients are suddenly made available.

physics.soc-ph↗

arXiv2013

Optimal Grouping for Group Minimax Hypothesis Testing

Bayesian hypothesis testing and minimax hypothesis testing represent extreme instances of detection in which the prior probabilities of the hypotheses are either completely and precisely known, or are completely unknown. Group minimax, also known as Gamma-minimax, is a robust intermediary between Bayesian and minimax hypothesis testing that allows for coarse or partial advance knowledge of the hypothesis priors by using information on sets in which the prior lies. Existing work on group minimax, however, does not consider the question of how to define the sets or groups of priors; it is assumed that the groups are given. In this work, we propose a novel intermediate detection scheme formulated through the quantization of the space of prior probabilities that optimally determines groups and also representative priors within the groups. We show that when viewed from a quantization perspective, group minimax amounts to determining centroids with a minimax Bayes risk error divergence distortion criterion: the appropriate Bregman divergence for this task. Moreover, the optimal partitioning of the space of prior probabilities is a Bregman Voronoi diagram. Together, the optimal grouping and representation points are an epsilon-net with respect to Bayes risk error divergence, and permit a rate-distortion type asymptotic analysis of detection performance with the number of groups. Examples of detecting signals corrupted by additive white Gaussian noise and of distinguishing exponentially-distributed signals are presented.

cs.IT↗

arXiv2012

Balancing Lifetime and Classification Accuracy of Wireless Sensor Networks

Wireless sensor networks are composed of distributed sensors that can be used for signal detection or classification. The likelihood functions of the hypotheses are often not known in advance, and decision rules have to be learned via supervised learning. A specific such algorithm is Fisher discriminant analysis (FDA), the classification accuracy of which has been previously studied in the context of wireless sensor networks. Previous work, however, does not take into account the communication protocol or battery lifetime of the sensor networks; in this paper we extend the existing studies by proposing a model that captures the relationship between battery lifetime and classification accuracy. In order to do so we combine the FDA with a model that captures the dynamics of the Carrier-Sense Multiple-Access (CSMA) algorithm, the random-access algorithm used to regulate communications in sensor networks. This allows us to study the interaction between the classification accuracy, battery lifetime and effort put towards learning, as well as the impact of the back-off rates of CSMA on the accuracy. We characterize the tradeoff between the length of the training stage and accuracy, and show that accuracy is non-monotone in the back-off rate due to changes in the training sample size and overfitting.

cs.NI↗

arXiv2012

Dynamic Matrix Factorization: A State Space Approach

Matrix factorization from a small number of observed entries has recently garnered much attention as the key ingredient of successful recommendation systems. One unresolved problem in this area is how to adapt current methods to handle changing user preferences over time. Recent proposals to address this issue are heuristic in nature and do not fully exploit the time-dependent structure of the problem. As a principled and general temporal formulation, we propose a dynamical state space model of matrix factorization. Our proposal builds upon probabilistic matrix factorization, a Bayesian model with Gaussian priors. We utilize results in state tracking, such as the Kalman filter, to provide accurate recommendations in the presence of both process and measurement noise. We show how system parameters can be learned via expectation-maximization and provide comparisons to current published techniques.

cs.LG↗

arXiv2008

Quantization of Prior Probabilities for Hypothesis Testing

Bayesian hypothesis testing is investigated when the prior probabilities of the hypotheses, taken as a random vector, are quantized. Nearest neighbor and centroid conditions are derived using mean Bayes risk error as a distortion measure for quantization. A high-resolution approximation to the distortion-rate function is also obtained. Human decision making in segregated populations is studied assuming Bayesian hypothesis testing with quantized priors.

cs.IT↗