Reinforcement Learning for Infinite-Horizon Average-Reward Linear MDPs via Approximation by Discounted-Reward MDPs

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Hong, Kihyuk, Chae, Woojin, Zhang, Yufan, Lee, Dabeen, Tewari, Ambuj
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912268910854144
author Hong, Kihyuk
Chae, Woojin
Zhang, Yufan
Lee, Dabeen
Tewari, Ambuj
author_facet Hong, Kihyuk
Chae, Woojin
Zhang, Yufan
Lee, Dabeen
Tewari, Ambuj
contents We study the problem of infinite-horizon average-reward reinforcement learning with linear Markov decision processes (MDPs). The associated Bellman operator of the problem not being a contraction makes the algorithm design challenging. Previous approaches either suffer from computational inefficiency or require strong assumptions on dynamics, such as ergodicity, for achieving a regret bound of $\widetilde{O}(\sqrt{T})$. In this paper, we propose the first algorithm that achieves $\widetilde{O}(\sqrt{T})$ regret with computational complexity polynomial in the problem parameters, without making strong assumptions on dynamics. Our approach approximates the average-reward setting by a discounted MDP with a carefully chosen discounting factor, and then applies an optimistic value iteration. We propose an algorithmic structure that plans for a nonstationary policy through optimistic value iteration and follows that policy until a specified information metric in the collected data doubles. Additionally, we introduce a value function clipping procedure for limiting the span of the value function for sample efficiency.
format Preprint
id arxiv_https___arxiv_org_abs_2405_15050
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Reinforcement Learning for Infinite-Horizon Average-Reward Linear MDPs via Approximation by Discounted-Reward MDPs
Hong, Kihyuk
Chae, Woojin
Zhang, Yufan
Lee, Dabeen
Tewari, Ambuj
Machine Learning
We study the problem of infinite-horizon average-reward reinforcement learning with linear Markov decision processes (MDPs). The associated Bellman operator of the problem not being a contraction makes the algorithm design challenging. Previous approaches either suffer from computational inefficiency or require strong assumptions on dynamics, such as ergodicity, for achieving a regret bound of $\widetilde{O}(\sqrt{T})$. In this paper, we propose the first algorithm that achieves $\widetilde{O}(\sqrt{T})$ regret with computational complexity polynomial in the problem parameters, without making strong assumptions on dynamics. Our approach approximates the average-reward setting by a discounted MDP with a carefully chosen discounting factor, and then applies an optimistic value iteration. We propose an algorithmic structure that plans for a nonstationary policy through optimistic value iteration and follows that policy until a specified information metric in the collected data doubles. Additionally, we introduce a value function clipping procedure for limiting the span of the value function for sample efficiency.
title Reinforcement Learning for Infinite-Horizon Average-Reward Linear MDPs via Approximation by Discounted-Reward MDPs
topic Machine Learning
url https://arxiv.org/abs/2405.15050