pith. sign in

arxiv: 2102.02981 · v2 · pith:GVY2YZ6Jnew · submitted 2021-02-05 · 💻 cs.LG · math.ST· stat.ML· stat.TH

Finite Sample Analysis of Minimax Offline Reinforcement Learning: Completeness, Fast Rates and First-Order Efficiency

classification 💻 cs.LG math.STstat.MLstat.TH
keywords completenessminimaxconvergenceefficiencyfastfirst-orderfunctionslearning
0
0 comments X
read the original abstract

We offer a theoretical characterization of off-policy evaluation (OPE) in reinforcement learning using function approximation for marginal importance weights and $q$-functions when these are estimated using recent minimax methods. Under various combinations of realizability and completeness assumptions, we show that the minimax approach enables us to achieve a fast rate of convergence for weights and quality functions, characterized by the critical inequality \citep{bartlett2005}. Based on this result, we analyze convergence rates for OPE. In particular, we introduce novel alternative completeness conditions under which OPE is feasible and we present the first finite-sample result with first-order efficiency in non-tabular environments, i.e., having the minimal coefficient in the leading term.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.