PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Tracking the best of many experts
Andras Gyorgy, Tamas Linder and Gábor Lugosi
In: COLT 2005, June 2005, Bertinoro, Italy.

Abstract

An algorithm is presented for online prediction that allows to track the best expert efficiently even if the number of experts is exponentially large, provided that the set of experts has a certain additive structure. As an example we work out the case when each expert is represented by a path in a directed graph and the loss of each expert is the sum of the weights over the edges in the path.

EPrint Type:Conference or Workshop Item (Paper)
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Computational, Information-Theoretic Learning with Statistics
Theory & Algorithms
ID Code:1363
Deposited By:Gábor Lugosi
Deposited On:28 November 2005