Toward theoretical understandings of robust Markov decision processes: sample complexity and asymptotics

DOI10.1214/22-AOS2225MaRDI QIDQ2112808

Zhihua Zhang, Liangyu Zhang, Wen Hao Yang

Publication date: 12 January 2023

Published in: The Annals of Statistics (Search for Journal in Brave)

Full work available at URL: https://arxiv.org/abs/2105.03863

zbMATH Keywords

distributional robustnessmodel-based reinforcement learning\(f\)-divergence setrobust MDPs

Mathematics Subject Classification ID

Asymptotic properties of parametric estimators (62F12) Computational learning theory (68Q32) General considerations in statistical decision theory (62C05)

Uses Software

• ElemStatLearn

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Doubly robust policy evaluation and optimization
• Minimax PAC bounds on the sample complexity of reinforcement learning with a generative model
• Learning models with uniform performance via distributionally robust optimization
• Recursive multiple-priors.
• Data-driven robust optimization
• A framework for optimization under ambiguity
• Robust Sensitivity Analysis for Stochastic Systems
• $f$ -Divergence Inequalities
• Distributionally Robust Optimization Under Moment Uncertainty with Application to Data-Driven Problems
• Asymptotic Statistics
• Distributionally Robust Stochastic Programming
• High-Dimensional Statistics
• Statistics of Robust Optimization: A Generalized Empirical Likelihood Approach
• Robust Markov Decision Processes
• Quantifying Distributional Model Risk via Optimal Transport
• Robust Control of Markov Decision Processes with Uncertain Transition Matrices
• Variance-based regularization with convex objectives
• Efficient Estimation of Average Treatment Effects Using the Estimated Propensity Score
• Robust Dynamic Programming

This page was built for publication: Toward theoretical understandings of robust Markov decision processes: sample complexity and asymptotics