Near-Optimal Offline Reinforcement Learning via Double Variance Reduction

by   Ming Yin, et al.

We consider the problem of offline reinforcement learning (RL) – a well-motivated setting of RL that aims at policy optimization using only historical data. Despite its wide applicability, theoretical understandings of offline RL, such as its optimal sample complexity, remain largely open even in basic settings such as tabular Markov Decision Processes (MDPs). In this paper, we propose Off-Policy Double Variance Reduction (OPDVR), a new variance reduction based algorithm for offline RL. Our main result shows that OPDVR provably identifies an ϵ-optimal policy with O(H^2/d_mϵ^2) episodes of offline data in the finite-horizon stationary transition setting, where H is the horizon length and d_m is the minimal marginal state-action distribution induced by the behavior policy. This improves over the best known upper bound by a factor of H. Moreover, we establish an information-theoretic lower bound of Ω(H^2/d_mϵ^2) which certifies that OPDVR is optimal up to logarithmic factors. Lastly, we show that OPDVR also achieves rate-optimal sample complexity under alternative settings such as the finite-horizon MDPs with non-stationary transitions and the infinite horizon MDPs with discounted rewards.



There are no comments yet.


page 1

page 2

page 3

page 4


Policy Finetuning: Bridging Sample-Efficient Offline and Online Reinforcement Learning

Recent theoretical work studies sample-efficient reinforcement learning ...

Towards Instance-Optimal Offline Reinforcement Learning with Pessimism

We study the offline reinforcement learning (offline RL) problem, where ...

Nearly Horizon-Free Offline Reinforcement Learning

We revisit offline reinforcement learning on episodic time-homogeneous t...

Bilinear Classes: A Structural Framework for Provable Generalization in RL

This work introduces Bilinear Classes, a new structural framework, which...

Is Pessimism Provably Efficient for Offline RL?

We study offline reinforcement learning (RL), which aims to learn an opt...

Optimal Uniform OPE and Model-based Offline Reinforcement Learning in Time-Homogeneous, Reward-Free and Task-Agnostic Settings

This work studies the statistical limits of uniform convergence for offl...

Bridging Offline Reinforcement Learning and Imitation Learning: A Tale of Pessimism

Offline (or batch) reinforcement learning (RL) algorithms seek to learn ...
This week in AI

Get the week's most popular data science and artificial intelligence research sent straight to your inbox every Saturday.