Policy Gradient for Robust Markov Decision Processes

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Wang, Qiuhao, Xu, Shaohang, Ho, Chin Pang, Petrik, Marek
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866915000432459776
author Wang, Qiuhao
Xu, Shaohang
Ho, Chin Pang
Petrik, Marek
author_facet Wang, Qiuhao
Xu, Shaohang
Ho, Chin Pang
Petrik, Marek
contents We develop a generic policy gradient method with the global optimality guarantee for robust Markov Decision Processes (MDPs). While policy gradient methods are widely used for solving dynamic decision problems due to their scalable and efficient nature, adapting these methods to account for model ambiguity has been challenging, often making it impractical to learn robust policies. This paper introduces a novel policy gradient method, Double-Loop Robust Policy Mirror Descent (DRPMD), for solving robust MDPs. DRPMD employs a general mirror descent update rule for the policy optimization with adaptive tolerance per iteration, guaranteeing convergence to a globally optimal policy. We provide a comprehensive analysis of DRPMD, including new convergence results under both direct and softmax parameterizations, and provide novel insights into the inner problem solution through Transition Mirror Ascent (TMA). Additionally, we propose innovative parametric transition kernels for both discrete and continuous state-action spaces, broadening the applicability of our approach. Empirical results validate the robustness and global convergence of DRPMD across various challenging robust MDP settings.
format Preprint
id arxiv_https___arxiv_org_abs_2410_22114
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Policy Gradient for Robust Markov Decision Processes
Wang, Qiuhao
Xu, Shaohang
Ho, Chin Pang
Petrik, Marek
Machine Learning
Artificial Intelligence
We develop a generic policy gradient method with the global optimality guarantee for robust Markov Decision Processes (MDPs). While policy gradient methods are widely used for solving dynamic decision problems due to their scalable and efficient nature, adapting these methods to account for model ambiguity has been challenging, often making it impractical to learn robust policies. This paper introduces a novel policy gradient method, Double-Loop Robust Policy Mirror Descent (DRPMD), for solving robust MDPs. DRPMD employs a general mirror descent update rule for the policy optimization with adaptive tolerance per iteration, guaranteeing convergence to a globally optimal policy. We provide a comprehensive analysis of DRPMD, including new convergence results under both direct and softmax parameterizations, and provide novel insights into the inner problem solution through Transition Mirror Ascent (TMA). Additionally, we propose innovative parametric transition kernels for both discrete and continuous state-action spaces, broadening the applicability of our approach. Empirical results validate the robustness and global convergence of DRPMD across various challenging robust MDP settings.
title Policy Gradient for Robust Markov Decision Processes
topic Machine Learning
Artificial Intelligence
url https://arxiv.org/abs/2410.22114