paper-with-me

Papers

Optimizing Percentile Criterion Using Robust MDPs

2019-10-23 · Bahram Behzadian, Reazul Hasan Russel, Marek Petrik, Chin Pang Ho

We address the problem of computing reliable policies in reinforcement learning problems with limited data. In particular, we compute policies that achieve good returns with high confidence when deployed. This objective, known as the \emph{percentile criterion}, can be optimized using Robust MDPs~(RMDPs). RMDPs generalize MDPs to allow for uncertain transition probabilities chosen adversarially from given ambiguity sets. We show that the RMDP solution's sub-optimality depends on the spans of the ambiguity sets along the value function. We then propose new algorithms that minimize the span of ambiguity sets defined by weighted $L_1$ and $L_\infty$ norms. Our primary focus is on Bayesian guarantees, but we also describe how our methods apply to frequentist guarantees and derive new concentration inequalities for weighted $L_1$ and $L_\infty$ norms. Experimental results indicate that our optimized ambiguity sets improve significantly on prior construction methods.

📄 PDF Abstract BibTeX arXiv:1910.10786

Code (0)

등록된 구현이 없습니다.

Tasks

Reinforcement LearningReinforcement Learning (RL)

Similar Papers 제목 키워드 기반

Percentile Criterion Optimization in Offline Reinforcement Learning

2024-04-07 · NeurIPS 2023 11 · Elita A. Lobo, Cyrus Cousins, Yair Zick, Marek Petrik

In reinforcement learning, robust policies for high-stakes decision-making problems with limited data are usually computed by optimizing the \emph{percentile criterion}. The percentile criterion is approximately solved b…

Decision Makingreinforcement-learningReinforcement Learning

Soft-Robust Algorithms for Batch Reinforcement Learning

2020-11-30 · Elita A. Lobo, Mohammad Ghavamzadeh, Marek Petrik

In reinforcement learning, robust policies for high-stakes decision-making problems with limited data are usually computed by optimizing the percentile criterion, which minimizes the probability of a catastrophic failure…

Decision Makingreinforcement-learningReinforcement LearningReinforcement Learning (RL)

Risk-Averse MDPs under Reward Ambiguity

2023-01-03 · Haolin Ruan, Zhi Chen, Chin Pang Ho

We propose a distributionally robust return-risk model for Markov decision processes (MDPs) under risk and reward ambiguity. The proposed model optimizes the weighted average of mean and percentile performances, and it c…

Risk-averse Total-reward MDPs with ERM and EVaR

2024-08-30 · Xihong Su, Julien Grand-Clément, Marek Petrik

Optimizing risk-averse objectives in discounted MDPs is challenging because most models do not admit direct dynamic programming equations and require complex history-dependent policies. In this paper, we show that the ri…

Optimizing Quantiles in Preference-based Markov Decision Processes

2016-12-01 · Hugo Gilbert, Paul Weng, Yan Xu

In the Markov decision process model, policies are usually evaluated by expected cumulative rewards. As this decision criterion is not always suitable, we propose in this paper an algorithm for computing a policy optimal…