A Unified View of TD Algorithms; Introducing Full-Gradient TD and Equi-Gradient Descent TD

dc.creatorLoth, Manuel
dc.creatorPreux, Philippe
dc.date2006-11-29
dc.date.accessioned2026-07-07T07:31:47Z
dc.date.available2026-07-07T07:31:47Z
dc.descriptionThis paper addresses the issue of policy evaluation in Markov Decision Processes, using linear function approximation. It provides a unified view of algorithms such as TD(lambda), LSTD(lambda), iLSTD, residual-gradient TD. It is asserted that they all consist in minimizing a gradient function and differ by the form of this function and their means of minimizing it. Two new schemes are introduced in that framework: Full-gradient TD which uses a generalization of the principle introduced in iLSTD, and EGD TD, which reduces the gradient by successive equi-gradient descents. These three algorithms form a new intermediate family with the interesting property of making much better use of the samples than TD while keeping a gradient descent scheme, which is useful for complexity issues and optimistic policy iteration.
dc.identifierhttps://arxiv.org/abs/cs/0611145
dc.identifierhttp://arxiv.org/abs/cs/0611145
dc.identifierDans European Symposium on Artificial Neural Networks (2006)
dc.identifier.urihttp://salesiana.dossiersoluciones.com/handle/123456789/118881
dc.subjectMachine Learning
dc.titleA Unified View of TD Algorithms; Introducing Full-Gradient TD and Equi-Gradient Descent TD
dc.typetext

Files

Collections