Gradient-based Reinforcement Planning in Policy-Search Methods

Simple item page
dc.creatorKwee, Ivo
dc.creatorHutter, Marcus
dc.creatorSchmidhuber, Juergen
dc.date2001-11-28
dc.date.accessioned2026-07-07T03:17:59Z
dc.date.available2026-07-07T03:17:59Z
dc.descriptionWe introduce a learning method called ``gradient-based reinforcement planning'' (GREP). Unlike traditional DP methods that improve their policy backwards in time, GREP is a gradient-based method that plans ahead and improves its policy before it actually acts in the environment. We derive formulas for the exact policy gradient that maximizes the expected future reward and confirm our ideas with numerical experiments.
dc.descriptionThis is an extended version of the paper presented at the EWRL 2001 in Utrecht (The Netherlands)
dc.identifierhttps://arxiv.org/abs/cs/0111060
dc.identifierhttp://arxiv.org/abs/cs/0111060
dc.identifier.urihttp://salesiana.dossiersoluciones.com/handle/123456789/30934
dc.subjectArtificial Intelligence
dc.subjectI.2; I.2.6; I.2.8
dc.titleGradient-based Reinforcement Planning in Policy-Search Methods
dc.typetext

Files

Collections

arXiv