We present a system for localizing sound sources in a room with severalad hocmicrophone arrays. Each circular array performs direction of arrival (DOA) estimation independently using commercial software. The DOAs are fed to a fusion center, concatenated, and used to perform the localization based on two proposed methods, which require only a few labeled source locations (anchor points) for training. The first proposed method is based on principal component analysis (PCA) of the observed DOA and does not require any knowledge of anchor points. The array cluster can then perform localization on a manifold defined by the PCA of concatenated DOAs over time. The second proposed method performs localization using an affine transformation between the DOA vectors and the room manifold. The PCA has fewer requirements on the training sequence, but is less robust to missing DOAs from one of the arrays. The methods are demonstrated with five IoT 8-microphone circular arrays, placed at unspecified fixed locations in an office. Both the PCA and the affine method can easily map out a rectangle based on a few anchor points with similar accuracy. The proposed methods provide a step toward monitoring activities in a smart home and require little installation effort as the array locations are not needed.

There has been a surge of works bridging MCMC sampling and optimization, with a specific focus on translating non-asymptotic convergence guarantees for optimization problems into the analysis of Langevin algorithms in MCMC sampling. A conspicuous distinction between the convergence analysis of Langevin sampling and that of optimization is that all known convergence rates for Langevin algorithms depend on the dimensionality of the problem, whereas the convergence rates for optimization are dimension-free for convex problems. Whether a dimension independent convergence rate can be achieved by the Langevin algorithm is thus a long-standing open problem. This paper provides an affirmative answer to this problem for the case of either Lipschitz or smooth convex functions with normal priors. By viewing Langevin algorithm as composite optimization, we develop a new analysis technique that leads to dimension independent convergence rates for such problems.

State-of-the-art implementations of boosting, such as XGBoost and LightGBM, can process large training sets extremely fast. However, this performance requires that the memory size is sufficient to hold a 2-3 multiple of the training set size. This paper presents an alternative approach to implementing the boosted trees, which achieves a significant speedup over XGBoost and LightGBM, especially when the memory size is small. This is achieved using a combination of three techniques: early stopping, effective sample size, and stratified sampling. Our experiments demonstrate a 10-100 speedup over XGBoost when the training data is too large to fit in memory.

We introduce a variant of the $k$-nearest neighbor classifier in which $k$ is chosen adaptively for each query, rather than supplied as a parameter. The choice of $k$ depends on properties of each neighborhood, and therefore may significantly vary between different points. (For example, the algorithm will use larger $k$ for predicting the labels of points in noisy regions.) We provide theory and experiments that demonstrate that the algorithm performs comparably to, and sometimes better than, $k$-NN with an optimal choice of $k$. In particular, we derive bounds on the convergence rates of our classifier that depend on a local quantity we call the ``advantage'' which is significantly weaker than the Lipschitz conditions used in previous convergence rate proofs. These generalization bounds hinge on a variant of the seminal Uniform Convergence Theorem due to Vapnik and Chervonenkis; this variant concerns conditional probabilities and may be of independent interest.

We argue that the second order bounds given in Cesa-Bianchi2006, which accumulate the square of the loss of each action separately, are loose. We propose a different form of a second order bound and conjecture the it is satisfied by NormalHedge ChaudhuriFrHs2009.

We address the problem of aggregating an ensemble of predictors with known loss bounds in a semi-supervised binary classification setting, to minimize prediction loss incurred on the unlabeled data. We find the minimax optimal predictions for a very general class of loss functions including all convex and many non-convex losses, extending a recent analysis of the problem for misclassification error. The result is a family of semi-supervised ensemble aggregation algorithms which are as efficient as linear learning by convex optimization, but are minimax optimal without any relaxations. Their decision rules take a form familiar in decision theory -- applying sigmoid functions to a notion of ensemble margin -- without the assumptions typically made in margin-based learning.

We develop a worst-case analysis of aggregation of classifier ensembles for binary classification. The task of predicting to minimize error is formulated as a game played over a given set of unlabeled data (a transductive setting), where prior label information is encoded as constraints on the game. The minimax solution of this game identifies cases where a weighted combination of the classifiers can perform significantly better than any single classifier.

We present and empirically evaluate an efficient algorithm that learns to aggregate the predictions of an ensemble of binary classifiers. The algorithm uses the structure of the ensemble predictions on unlabeled data to yield significant performance improvements. It does this without making assumptions on the structure or origin of the ensemble, without parameters, and as scalably as linear learning. We empirically demonstrate these performance gains with random forests.

Robust real time tracking is a requirement for many emerging applications. Many of these applications must track objects even as their appearance changes. Training classifiers online has become an effective approach for dealing with variability in object appearance. Classifiers can learn and adapt to changes online at the cost of additional runtime computation. In this paper, we propose a FPGA accelerated design of an online boosting algorithm that uses multiple classifiers to track and recover objects in real time. Our algorithm uses a novel method for training and comparing pose-specific classifiers along with adaptive tracking classifiers. Our FPGA accelerated design is able to track at 60 frames per second while concurrently evaluating 11 classifiers. This represents a 30× speed up over a CPU based software implementation. It also demonstrates tracking accuracy at state of the art levels on a standard set of videos.

The traditional k-NN classification rule predicts a label based on the most common label of the k nearest neighbors (the plurality rule). It is known that the plurality rule is optimal when the number of examples tends to infinity. In this paper we show that the plurality rule is sub-optimal when the number of labels is large and the number of examples is small. We propose a simple k-NN rule that takes into account the labels of all of the neighbors, rather than just the most common label. We present a number of experiments on both synthetic datasets and real-world datasets, including MNIST and SVHN. We show that our new rule can achieve lower error rates compared to the majority rule in many cases.

Hints are sometimes used in online learning system to help students when they are having difficulties. However, in all of the systems we are aware of, the hints are fixed ahead of time and do not depend on the unsuccessful attempts the student has already made. This severely limits the effectiveness of the hints.

We prove the first finite-sample convergence rates for any incremental PCA algorithm using sub-quadratic time and memory per iteration. The algorithm analyzed is Oja's learning rule, an efficient and well-known scheme for estimating the top principal component. Our analysis of this non-convex problem yields expected and high-probability convergence rates of $\tilde{O}(1/n)$ through a novel technique. We relate our guarantees to existing rates for stochastic gradient descent on strongly convex functions, and extend those results. We also include experiments which demonstrate convergence behaviors predicted by our analysis.

We present RIFFA, a reusable integration framework for FPGA accelerators. RIFFA provides communication and synchronization for FPGA accelerated software using a standard interface. Our goal is to expand the use of FPGAs as an acceleration platform by releasing, as open source, a no cost framework that easily integrates software on traditional CPUs with FPGA based IP cores, over PCIe, with minimal custom configuration. RIFFA requires no specialized hardware or fee licensed IP cores. It can be deployed on common Linux workstations with a PCIe bus and has been tested on two different Linux distributions using Xilinx FPGAs.

We propose a novel online learning-based framework for occlusion boundary detection in video sequences. This approach does not require any prior training and instead "learns" occlusion boundaries by updating a set of weights for the online learning Hedge algorithm at each frame instance. Whereas previous training-based methods perform well only on data similar to the trained examples, the proposed method is well suited for any video sequence. We demonstrate the performance of the proposed detector both for the CMU data set, which includes hand-labeled occlusion boundaries, and for a novel video sequence. In addition to occlusion boundary detection, the proposed algorithm is capable of classifying occlusion boundaries by angle and by whether the occluding object is covering or uncovering the background.

In this work, a novel occlusion detection algorithm using online learning is proposed for video applications. Each frame of a video is considered as a time-step for which pixels are classified as being either occluded or non-occluded. The Hedge algorithm is employed to determine weights for a set of experts, each of which is tuned to detect a specific type of occlusion boundary. In contrast to previous training-based methods, the proposed algorithm does not require any training, and has a runtime linear with respect to the number of experts considered. Detection performance is excellent on novel video sequences for which training data does not exist. In addition, the proposed algorithm is easily extended to provide classification results supplementary to detection. We demonstrate results on a series of challenging video sequences including a dataset of hand-labelled occlusion boundaries.

The Fluidic Lens Camera System provides an exciting opportunity for the Image Processing Community. Designed for a surgical environment, this camera has higher magnification and has better portability than traditional laparoscopic cameras. From an image processing prospective, the fluid causes non-uniform blur of different color planes. While the green image is sharp, the red and blue images are blurred. Previous methods have been developed to separate out the edge and shading components of the green image and to use the edge information in green to replace the blurred blue edges. This algorithm succeed in most areas, however in some areas, color bleeding artifacts occurred. We restate this problem as a classification problem. Using the contourlet and wavelet coefficients as features, the proposed algorithm determines in what areas color bleeding will occur and does not apply the sharpening algorithm in these areas. By applying the previous contourlet method in areas where it succeeds, we can produce an overall sharper image with reduced color bleeding artifacts. The ability to correctly classify when the previous algorithm will succeed is crucial to the success of the algorithm. The principal application is medical imaging, however, the fields of satellite pan-sharpening and image denoising can benefit from the results found in this paper.

Massive quantities of digital data are being collected in every aspect of modern life. Examples include Personal photos and videos, biological and medical images and recordings from sensor arrays. To transform these massive data streams into useful information we use a sequence of "winnowing" stages. Each step reduces the size of the data by an order of magnitude; extracting the wheat form the chaff. In this talk I will describe this approach in a variety of contexts, ranging from the analysis of genetic pathways in fruit-fly embryos and C-Elegans worms to counting birds and helping elderly people living alone keep in touch with their family and caregivers.

We have built a system that engages naive users in an audio-visual interaction with a computer in an unconstrained public space. We combine audio source localization techniques with face detection algorithms to detect and track the user throughout a large lobby. The sensors we use are an ad-hoc microphone array and a PTZ camera. To engage the user, the PTZ camera turns and points at sounds made by people passing by. From this simple pointing of a camera, the user is made aware that the system has acknowledged their presence. To further engage the user, we develop a face classification method that identifies and then greets previously seen users. The user can interact with the system through a simple hot-spot based gesture interface. To make the user interactions with the system feel natural, we utilize reconfigurable hardware, achieving a visual response time of less than 100ms. We rely heavily on machine learning methods to make our system self-calibrating and adaptive.

No abstract available.

We study the problem of decision-theoretic online learning (DTOL). Motivated by practical applications, we focus on DTOL when the number of actions is very large. Previous algorithms for learning in this framework have a tunable learning rate parameter, and a major barrier to using online-learning in practical applications is that it is not understood how to set this parameter optimally, particularly when the number of actions is large. In this paper, we offer a clean solution by proposing a novel and completely parameter-free algorithm for DTOL. In addition, we introduce a new notion of regret, which is more natural for applications with a large number of actions. We show that our algorithm achieves good performance with respect to this new notion of regret; in addition, it also achieves performance close to that of the best bounds achieved by previous algorithms with optimally-tuned parameters, according to previous notions of regret.

BACKGROUND: Identifying the catalytic residues in enzymes can aid in understanding the molecular basis of an enzyme's function and has significant implications for designing new drugs, identifying genetic disorders, and engineering proteins with novel functions. Since experimentally determining catalytic sites is expensive, better computational methods for identifying catalytic residues are needed. RESULTS: We propose ResBoost, a new computational method to learn characteristics of catalytic residues. The method effectively selects and combines rules of thumb into a simple, easily interpretable logical expression that can be used for prediction. We formally define the rules of thumb that are often used to narrow the list of candidate residues, including residue evolutionary conservation, 3D clustering, solvent accessibility, and hydrophilicity. ResBoost builds on two methods from machine learning, the AdaBoost algorithm and Alternating Decision Trees, and provides precise control over the inherent trade-off between sensitivity and specificity. We evaluated ResBoost using cross-validation on a dataset of 100 enzymes from the hand-curated Catalytic Site Atlas (CSA). CONCLUSION: ResBoost achieved 85% sensitivity for a 9.8% false positive rate and 73% sensitivity for a 5.7% false positive rate. ResBoost reduces the number of false positives by up to 56% compared to the use of evolutionary conservation scoring alone. We also illustrate the ability of ResBoost to identify recently validated catalytic residues not listed in the CSA.

A simple and computationally efficient scheme for tree-structured vector quantization is presented. Unlike previous methods, its quantization error depends only on the intrinsic dimension of the data distribution, rather than the apparent dimension of the space in which the data happen to lie.

We present a simple variant of the k-d tree which automatically adapts to intrinsic low dimensional structure in data without having to explicitly learn this structure.

We present a simple variant of the k-d tree which automatically adapts to intrinsic low dimensional structure in data.

BACKGROUND: Existing large-scale metabolic models of sequenced organisms commonly include enzymatic functions which can not be attributed to any gene in that organism. Existing computational strategies for identifying such missing genes rely primarily on sequence homology to known enzyme-encoding genes. RESULTS: We present a novel method for identifying genes encoding for a specific metabolic function based on a local structure of metabolic network and multiple types of functional association evidence, including clustering of genes on the chromosome, similarity of phylogenetic profiles, gene expression, protein fusion events and others. Using E. coli and S. cerevisiae metabolic networks, we illustrate predictive ability of each individual type of association evidence and show that significantly better predictions can be obtained based on the combination of all data. In this way our method is able to predict 60% of enzyme-encoding genes of E. coli metabolism within the top 10 (out of 3551) candidates for their enzymatic function, and as a top candidate within 43% of the cases. CONCLUSION: We illustrate that a combination of genome context and other functional association evidence is effective in predicting genes encoding metabolic enzymes. Our approach does not rely on direct sequence homology to known enzyme-encoding genes, and can be used in conjunction with traditional homology-based metabolic reconstruction methods. The method can also be used to target orphan metabolic activities.

BACKGROUND: We have recently introduced a predictive framework for studying gene transcriptional regulation in simpler organisms using a novel supervised learning algorithm called GeneClass. GeneClass is motivated by the hypothesis that in model organisms such as Saccharomyces cerevisiae, we can learn a decision rule for predicting whether a gene is up- or down-regulated in a particular microarray experiment based on the presence of binding site subsequences ("motifs") in the gene's regulatory region and the expression levels of regulators such as transcription factors in the experiment ("parents"). GeneClass formulates the learning task as a classification problem--predicting +1 and -1 labels corresponding to up- and down-regulation beyond the levels of biological and measurement noise in microarray measurements. Using the Adaboost algorithm, GeneClass learns a prediction function in the form of an alternating decision tree, a margin-based generalization of a decision tree. METHODS: In the current work, we introduce a new, robust version of the GeneClass algorithm that increases stability and computational efficiency, yielding a more scalable and reliable predictive model. The improved stability of the prediction tree enables us to introduce a detailed post-processing framework for biological interpretation, including individual and group target gene analysis to reveal condition-specific regulation programs and to suggest signaling pathways. Robust GeneClass uses a novel stabilized variant of boosting that allows a set of correlated features, rather than single features, to be included at nodes of the tree; in this way, biologically important features that are correlated with the single best feature are retained rather than decorrelated and lost in the next round of boosting. Other computational developments include fast matrix computation of the loss function for all features, allowing scalability to large datasets, and the use of abstaining weak rules, which results in a more shallow and interpretable tree. We also show how to incorporate genome-wide protein-DNA binding data from ChIP chip experiments into the GeneClass algorithm, and we use an improved noise model for gene expression data. RESULTS: Using the improved scalability of Robust GeneClass, we present larger scale experiments on a yeast environmental stress dataset, training and testing on all genes and using a comprehensive set of potential regulators. We demonstrate the improved stability of the features in the learned prediction tree, and we show the utility of the post-processing framework by analyzing two groups of genes in yeast--the protein chaperones and a set of putative targets of the Nrg1 and Nrg2 transcription factors--and suggesting novel hypotheses about their transcriptional and post-transcriptional regulation. Detailed results and Robust GeneClass source code is available for download from http://www.cs.columbia.edu/compbio/robust-geneclass.

One significant challenge in the construction of visual detection systems is the acquisition of sufficient labeled data. We describe a new technique for training visual detectors which requires only a small quantity of labeled data, and then uses unlabeled data to improve performance over time. Unsupervised improvement is based on the cotraining framework of Blum and Mitchell, in which two disparate classifiers are trained simultaneously. Unlabeled examples which are confidently labeled by one classifier are added, with labels, to the training set of the other classifier. Experiments are presented on the realistic task of automobile detection in roadway surveillance video. In this application, cotraining reduces the false positive rate by a factor of 2 to 11 from the classifier trained with labeled data alone.

In the multiarmed bandit problem, a gambler must decide which arm of K nonidentical slot machines to play in a sequence of trials so as to maximize his reward. This classical problem has received much attention because of the simple model it provides of the trade-off between exploration (trying out each arm to find the best one) and exploitation (playing the arm believed to give the best payoff). Past solutions for the bandit problem have almost always relied on assumptions about the statistics of the slot machines. In this work, we make no statistical assumptions whatsoever about the nature of the process generating the payoffs of the slot machines. We give a solution to the bandit problem in which an adversary, rather than a well-behaved stochastic process, has complete control over the payoffs. In a sequence of T plays, we prove that the per-round payoff of our algorithm approaches that of the best arm at the rate O(T -1/2 ). We show by a matching lower bound that this is the best possible. We also prove that our algorithm approaches the per-round payoff of any set of strategies at a similar rate: if the best strategy is chosen from a pool of N strategies, then our algorithm approaches the per-round payoff of the strategy at the rate O((log N 1/2 T -1/2 ). Finally, we apply our results to the problem of playing an unknown repeated matrix game. We show that our algorithm approaches the minimax payoff of the unknown game at the rate O(T -1/2 ).

We propose a new boosting algorithm. This boosting algorithm is an adaptive version of the boost by majority algorithm and combines bounded goals of the boost by majority algorithm with the adaptivity of AdaBoost. The method used for making boost-by-majority adaptive is to consider the limit in which each of the boosting iterations makes an infinitesimally small contribution to the process as a whole. This limit can be modeled using the differential equations that govern Brownian motion. The new boosting algorithm, named BrownBoost, is based on finding solutions to these differential equations. The paper describes two methods for finding approximate solutions to the differential equations. The first is a method that results in a provably polynomial time algorithm. The second method, based on the Newton-Raphson minimization procedure, is much more efficient in practice but is not known to be polynomial. 1

Article Estimating a mixture of two product distributions Share on Authors: Yoav Freund AT&T Labs, 180 Park Avenue, Florham Park, NJ AT&T Labs, 180 Park Avenue, Florham Park, NJView Profile , Yishay Mansour AT&T Labs and Tel-Aviv University AT&T Labs and Tel-Aviv UniversityView Profile Authors Info & Claims COLT '99: Proceedings of the twelfth annual conference on Computational learning theoryJuly 1999 Pages 53–62https://doi.org/10.1145/307400.307412Online:06 July 1999Publication History 26citation351DownloadsMetricsTotal Citations26Total Downloads351Last 12 Months2Last 6 weeks0 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my AlertsNew Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteGet Access

Most of the work which attempts to give bounds on the generalization error of the hypothesis generated by a learning algorithm is based on methods from the theory of uniform convergence. These bounds are a-priori bounds that hold for any distribution of examples and are calculated before any data is observed. In this paper we propose a different approach for bounding the generalization error after the data has been observed. A self-bounding learning algorithm is an algorithm which, in addition to the hypothesis that it outputs, outputs a reliable upper bound on the generalization error of this hypothesis. We first explore the idea in the statistical query learning framework of Kearns [10]. After that we give an explicit self bounding algorithm for learning algorithms that are based on local search. 1 INTRODUCTION Most of the work on the sample complexity of learning is based on uniform convergence theory and attempts to give uniform a-priori bounds. A uniform a-priori bound is a guar...

We introduce and analyze a new algorithm for linear classification which combines Rosenblatt 's perceptron algorithm with Helmbold and Warmuth's leave-one-out method. Like Vapnik 's maximal-margin classifier, our algorithm takes advantage of data that are linearly separable with large margins. Compared to Vapnik's algorithm, however, ours is much simpler to implement, and much more efficient in terms of computation time. We also show that our algorithm can be efficiently used in very high dimensional spaces using kernel functions. We performed some experiments using our algorithm, and some variants of it, for classifying images of handwritten digits. The performance of our algorithm is close to, but not as good as, the performance of maximal-margin classifiers on the same problem, while saving significantly on computation time and programming effort. 1 Introduction One of the most influential developments in the theory of machine learning in the last few years is Vapnik's work on supp...

We study online learning algorithms that predict by combining the predictions of severrd subordinate prediction algorithms, sometimes crdled "experts ."These simple algorithms belong to the multiplicative weights family of algorithms.The performance of these algorithms degrades only logarithmically with the number of experts, making them particularly useful in applications where the number of experts is very large.However, in applications such as text categorization, it is often natural for some of the experts to abstain from making predictions on some of the instances.We show how to transform algorithms that assume that afl experts are atways awake to algorithms that do not require this assumption.We also show how to derive corresponding Ioss bounds.Our method is very generaf, and can be applied to a large family of online learning algori[hms.We also give applications to various prediction models including decision graphs and "switching" experts.

This paper describes new and efficient algorithms for learning deterministic finite automata. Our approach is primarily distinguished by two features: (1) the adoption of an average-case setting to model the “typical” labeling of a finite automaton, while retaining a worst-case model for the underlying graph of the automaton, along with (2) a learning model in which the learner is not provided with the means to experiment with the machine, but rather must learn solely by observing the automaton's output behavior on a random input sequence. The main contribution of this paper is in presenting the first efficient algorithms for learning non-trivial classes of automata in an entirely passive learning model. We adopt an on-line learning model in which the learner is asked to predict the output of the next state, given the next symbol of the random input sequence; the goal of the learner is to make as few prediction mistakes as possible. Assuming the learner has a means of resetting the target machine to a fixed start state, we first present an efficient algorithm that makes an expected polynomial number of mistakes in this model. Next, we show how this first algorithm can be used as a subroutine by a second algorithm that also makes a polynomial number of mistakes even in the absence of a reset. Along the way, we prove a number of combinatorial results for randomly labeled automata. We also show that the labeling of the states and the bits of the input sequence need not be truly random, but merely semi - random . Finally, we discuss an extension of our results to a model in which automata are used to represent distributions over binary strings.

We analyze algorithms that predict a binary value by combining the predictions of several prediction strategies, calledexperts. Our analysis is for worst-case situations, i.e., we make no assumptions about the way the sequence of bits to be predicted is generated. We measure the performance of the algorithm by the difference between the expected number of mistakes it makes on the bit sequence and the expected number of mistakes made by the best expert on this sequence, where the expectation is taken with respect to the randomization in the predictins. We show that the minimum achievable difference is on the order of the square root of the number of mistakes of the best expert, and we give efficient algorithms that achieve this. Our upper and lower bounds have matching leading constants in most cases. We then show how this leads to certain kinds of pattern recognition/learning algorithms with performance bounds that improve on the best results currently know in this context. We also compare our analysis to the case in which log loss is used instead of the expected number of mistakes.

We apply the exponential weight algorithm, Bayes algorithm and is better than it for prediction in the worst-case.

We study the close connections between game theory, on-line prediction and boosting.After a brief review of game theory, we describe an algorithm for learning to play repeated games based on the on-line prediction methods of Littlestone and Warmuth.The analysis of this algorithm yields a simple proof of von Neumann's famous minmax theorem, as well as a provable method of approximately solving a game.We then show that the on-line prediction model is obtained by applying this gameplaying algorithm to an appropriate choice of game and that boosting is obtained by applying the same algorithm to the "dual" of this game.

We study efficient algorithms for solving the following problem, which we call the switching distributions learning problem.A sequence S = alaz... an, over a finite alphabet Z is generated in the following way.The sequence is a concatenation of K runs, each of which is a consecutive subsequence.Each run is generated by independent random draws from a distribution 17i over Z, where $% is an element in a set of distributions {p,,..., f?~}.The learning algorithm is given this sequence and its goal is to find approximations of the distributions ~1, . . . .$IV, and give an approximate segmentation of the sequence into its constituting runs.We give an efficient algorithm for solving this problem and show conditions under which the algorithm is guaranteed to work with high probability.

In the multi-armed bandit problem, a gambler must decide which arm of K non-identical slot machines to play in a sequence of trials so as to maximize his reward. This classical problem has received much attention because of the simple model it provides of the trade-off between exploration (trying out each arm to find the best one) and exploitation (playing the arm believed to give the best payoff). Past solutions for the bandit problem have almost always relied on assumptions about the statistics of the slot machines. In this work, we make no statistical assumptions whatsoever about the nature of the process generating the payoffs of the slot machines. We give a solution to the bandit problem in which an adversary, rather than a well-behaved stochastic process, has complete control over the payoffs. In a sequence of T plays, we prove that the expected per-round payoff of our algorithm approaches that of the best arm at the rate O(T/sup -1/3/), and we give an improved rate of convergence when the best arm has fairly low payoff. We also consider a setting in which the player has a team of "experts" advising him on which arm to play; here, we give a strategy that will guarantee expected payoff close to that of the best expert. Finally, we apply our result to the problem of learning to play an unknown repeated matrix game against an all-powerful adversary.

We examine the problem of learning to play various games optimally against resource-bounded adversaries, with an explicit emphasis on the computational efficiency of the learning algorithm. We are especially interested in providing efficient algorithms for games other than penny-matching (in which payoff is received for matching the adversary's action in the current round), and for adversaries other than the classically studied finite automata. In particular, we examine games and adversaries for which the learning algorithm's past actions may strongly affect the adversary's future willingness to "cooperate" (that is, permit high payoff), and therefore require carefully planned actions on the part of the learning algorithm. For example, in the game we call contract, both sides play O or 1 on each round, but our side receives payoff only if we play 1 in synchrony with the adversary; unlike penny-matching, playing O in synchrony with the adversary pays nothing. The name of the game is derived from the example of signing a contract, which becomes valid only if both parties sign (play 1).

Article How to use expert advice Share on Authors: Nicolò Cesa-Bianchi View Profile , Yoav Freund View Profile , David P. Helmbold View Profile , David Haussler View Profile , Robert E. Schapire View Profile , Manfred K. Warmuth View Profile Authors Info & Claims STOC '93: Proceedings of the twenty-fifth annual ACM symposium on Theory of ComputingJune 1993 Pages 382–391https://doi.org/10.1145/167088.167198Online:01 June 1993Publication History 71citation406DownloadsMetricsTotal Citations71Total Downloads406Last 12 Months8Last 6 weeks3 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my AlertsNew Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteGet Access

Article Efficient learning of typical finite automata from random walks Share on Authors: Yoav Freund View Profile , Michael Kearns View Profile , Dana Ron View Profile , Ronitt Rubinfeld View Profile , Robert E. Schapire View Profile , Linda Sellie View Profile Authors Info & Claims STOC '93: Proceedings of the twenty-fifth annual ACM symposium on Theory of ComputingJune 1993 Pages 315–324https://doi.org/10.1145/167088.167191Published:01 June 1993 28citation470DownloadsMetricsTotal Citations28Total Downloads470Last 12 Months3Last 6 weeks0 Get Citation AlertsNew Citation Alert added!This alert has been successfully added and will be sent to:You will be notified whenever a record that you have chosen has been cited.To manage your alert preferences, click on the button below.Manage my AlertsNew Citation Alert!Please log in to your account Save to BinderSave to BinderCreate a New BinderNameCancelCreateExport CitationPublisher SiteGet Access

In this work we present some improvements and extensions to previous work on boosting weak learners [Sch90, Fre90]. Our main result is an improvement of the boosting-by-majority algorithm. One implication of the performance of this algorithm is that if a concept class can be learned in the PAC model to within some fixed error smaller than 1/2, then it can be learned to within an arbitrarily small error ε > 0 with time complexity 0((1/ε)(log 1/ε)2) (fixing the sample space and concept class and the required reliability). We show that the majority rule is the optimal rule for combining general weak learners. We also extend the boosting algorithm to concept classes that give multi-valued labels and real-valued labels.