A desicion-theoretic generalization of on-line learning and an application to boosting
We consider the problem of dynamically apportioning resources among a set of options in a worst-case on-line framework. The model we study can be interpreted as a broad, abstract extension of the well-studied on-line prediction model to a general decision-theoretic setting. We show that the multiplicative weight-update rule of Littlestone and Warmuth  can be adapted to this mode yielding bounds that are slightly weaker in some cases, but applicable to a considerably more general class of learning problems. We show how the resulting learning algorithm can be applied to a variety of problems, including gambling, multiple-outcome prediction, repeated games and prediction of points in ℝ n . We also show how the weight-update rule can be used to derive a new boosting algorithm which does not require prior knowledge about the performance of the weak learning algorithm.
KeywordsLoss Function Weak Hypothesis Algorithm AdaBoost Final Hypothesis Cumulative Loss
Unable to display preview. Download preview PDF.
- 1.Nicolò Cesa-Bianchi, Yoav Freund, David P. Helmbold, David Haussler, Robert E. Schapire, and Manfred K. Warmuth. How to use expert advice. In Proceedings of the Twenty-Fifth Annual ACM Symposium on the Theory of Computing, pages 382–391, May 1993.Google Scholar
- 2.Thomas H. Chung. Approximate methods for sequential decision making using expert advice. In Proceedings of the Seventh Annual ACM Conference on Computational Learning Theory, pages 183–189, 1994.Google Scholar
- 3.Thomas M. Cover. Universal portfolios. Mathematics of Finance, 1(1):1–29, January 1991.Google Scholar
- 4.Harris Drucker, Robert Schapire, and Patrice Simard. Boosting performance in neural networks. International Journal of Pattern Recognition and Artificial Intelligence, 7(4):705–719, 1993.Google Scholar
- 5.Yoav Freund. Data Filtering and Distribution Modeling Algorithms for Machine Learning. PhD thesis, University of California at Santa Cruz, 1993. Retrievable from: ftp.cse.ucsc.edu/pub/tr/ucsc-crl-93-37.ps.Z.Google Scholar
- 6.Yoav Freund. Boosting a weak learning algorithm by majority. Information and Computation, To appear.Google Scholar
- 7.James Hannan. Approximation to Bayes risk in repeated play. In M. Dresher, A. W. Tucker, and P. Wolfe, editors, Contributions to the Theory of Games, volume III, pages 97–139. Princeton University Press, 1957.Google Scholar
- 8.David Haussler, Jyrki Kivinen, and Manfred K. Warmuth. Tight worst-case loss bounds for predicting with expert advice. In Proceedings of the Second European Conference on Computational Learning Theory, 1995.Google Scholar
- 9.Jyrki Kivinen and Manfred K. Warmuth. Using experts for predicting continuous outcomes. In Computational Learning Theory: EuroCOLT'93, pages 109–120, 1994.Google Scholar
- 10.Nick Littlestone and Manfred K. Warmuth. The weighted majority algorithm. Information and Computation, 108:212–261, 1994.Google Scholar
- 11.Robert E. Schapire. The strength of weak learnability. Machine Learning, 5(2):197–227, 1990.Google Scholar
- 12.Volodimir G. Vovk. Aggregating strategies. In Proceedings of the Third Annual Workshop on Computational Learning Theory, pages 371–383, 1990.Google Scholar