A Primal-Dual Algorithm for Offline Constrained Reinforcement Learning with Linear MDPs

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Hong, Kihyuk, Tewari, Ambuj
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866916270391164928
author Hong, Kihyuk
Tewari, Ambuj
author_facet Hong, Kihyuk
Tewari, Ambuj
contents We study offline reinforcement learning (RL) with linear MDPs under the infinite-horizon discounted setting which aims to learn a policy that maximizes the expected discounted cumulative reward using a pre-collected dataset. Existing algorithms for this setting either require a uniform data coverage assumptions or are computationally inefficient for finding an $ε$-optimal policy with $O(ε^{-2})$ sample complexity. In this paper, we propose a primal dual algorithm for offline RL with linear MDPs in the infinite-horizon discounted setting. Our algorithm is the first computationally efficient algorithm in this setting that achieves sample complexity of $O(ε^{-2})$ with partial data coverage assumption. Our work is an improvement upon a recent work that requires $O(ε^{-4})$ samples. Moreover, we extend our algorithm to work in the offline constrained RL setting that enforces constraints on additional reward signals.
format Preprint
id arxiv_https___arxiv_org_abs_2402_04493
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle A Primal-Dual Algorithm for Offline Constrained Reinforcement Learning with Linear MDPs
Hong, Kihyuk
Tewari, Ambuj
Machine Learning
We study offline reinforcement learning (RL) with linear MDPs under the infinite-horizon discounted setting which aims to learn a policy that maximizes the expected discounted cumulative reward using a pre-collected dataset. Existing algorithms for this setting either require a uniform data coverage assumptions or are computationally inefficient for finding an $ε$-optimal policy with $O(ε^{-2})$ sample complexity. In this paper, we propose a primal dual algorithm for offline RL with linear MDPs in the infinite-horizon discounted setting. Our algorithm is the first computationally efficient algorithm in this setting that achieves sample complexity of $O(ε^{-2})$ with partial data coverage assumption. Our work is an improvement upon a recent work that requires $O(ε^{-4})$ samples. Moreover, we extend our algorithm to work in the offline constrained RL setting that enforces constraints on additional reward signals.
title A Primal-Dual Algorithm for Offline Constrained Reinforcement Learning with Linear MDPs
topic Machine Learning
url https://arxiv.org/abs/2402.04493