A new Hedging algorithm and its application to inferring latent random variables

Simple item page
dc.creatorFreund, Yoav
dc.creatorHsu, Daniel
dc.date2008-06-30
dc.date.accessioned2026-07-07T09:47:25Z
dc.date.available2026-07-07T09:47:25Z
dc.descriptionWe present a new online learning algorithm for cumulative discounted gain. This learning algorithm does not use exponential weights on the experts. Instead, it uses a weighting scheme that depends on the regret of the master algorithm relative to the experts. In particular, experts whose discounted cumulative gain is smaller (worse) than that of the master algorithm receive zero weight. We also sketch how a regret-based algorithm can be used as an alternative to Bayesian averaging in the context of inferring latent random variables.
dc.identifierhttps://arxiv.org/abs/0806.4802
dc.identifierhttp://arxiv.org/abs/0806.4802
dc.identifier.urihttp://salesiana.dossiersoluciones.com/handle/123456789/163870
dc.subjectComputer Science and Game Theory
dc.subjectArtificial Intelligence
dc.titleA new Hedging algorithm and its application to inferring latent random variables
dc.typetext

Files

Collections

arXiv