Off-policy Learning with Eligibility Traces: A Survey
Abstract
In the framework of Markov Decision Processes, we consider linear \emph{off-policy} learning, that is the problem of learning a linear approximation of the value function of some fixed policy from one trajectory possibly generated by some other policy. We briefly review \emph{on-policy} learning algorithms of the literature (gradient-based and least-squares-based), adopting a unified algorithmic view. Then, we highlight a systematic approach for adapting them to \emph{off-policy} learning \emph{with eligibility traces}. This leads to some known algorithms -- off-policy LSTD($\lambda$), LSPE($\lambda$), TD($\lambda$), TDC/GQ($\lambda$) -- and suggests new extensions -- off-policy FPKF($\lambda$), BRM($\lambda$), gBRM($\lambda$), GTD2($\lambda$). We describe a comprehensive algorithmic derivation of all algorithms in a recursive and memory-efficent form, discuss their known convergence properties and illustrate their relative empirical behavior on Garnet problems. Our experiments suggest that the most standard algorithms on and off-policy LSTD($\lambda$)/LSPE($\lambda$) -- and TD($\lambda$) if the feature space dimension is too large for a least-squares approach -- perform the best.
Origin | Files produced by the author(s) |
---|