Prediction, Learning, and Games (Google eBook)

Front Cover
Cambridge University Press, Mar 13, 2006 - Computers
0 Reviews
This important new text and reference for researchers and students in machine learning, game theory, statistics and information theory offers the first comprehensive treatment of the problem of predicting individual sequences. Unlike standard statistical approaches to forecasting, prediction of individual sequences does not impose any probabilistic assumption on the data-generating mechanism. Yet, prediction algorithms can be constructed that work well for all possible sequences, in the sense that their performance is always nearly as good as the best forecasting strategy in a given reference class. The central theme is the model of prediction using expert advice, a general framework within which many related problems can be cast and discussed. Repeated game playing, adaptive data compression, sequential investment in the stock market, sequential pattern analysis, and several other problems are viewed as instances of the experts' framework and analyzed from a common nonstochastic standpoint that often reveals new and intriguing connections. Old and new forecasting methods are described in a mathematically precise way in order to characterize their theoretical limitations and possibilities.
  

What people are saying - Write a review

We haven't found any reviews in the usual places.

Contents

Preface page xi
1
Prediction with Expert Advice
7
Tight Bounds for Specific Losses
40
Randomized Prediction
67
Efficient Forecasters for Large Classes of Experts
99
Prediction with Limited Feedback
128
Prediction and Playing Games
180
Absolute Loss
233
Logarithmic Loss
247
Sequential Investment
276
Linear Pattern Recognition
293
Linear Classification
333
References
373
Author Index 387
390
Copyright

Common terms and phrases

References to this book

All Book Search results »

About the author (2006)

Nicolň Cesa-Bianchi is Professor of Computer Science at the University of Milan, Italy. His research interests include learning theory, pattern analysis, and worst-case analysis of algorithms. He is the acting editor of The Machine Learning Journal.

Gábor Lugosi has been working on various problems in pattern classification, nonparametric statistics, statistical learning theory, game theory, probability, and information theory. He is co-author of the monographs, A Probabilistic Theory of Pattern Recognition and Combinatorial Methods of Density Estimation. He has been an associate editor of various journals including The IEEE Transactions of Information Theory, Test, ESAIM: Probability and Statistics and Statistics and Decisions.

Bibliographic information