paper-with-me

홈 › Papers

Policy Certificates: Towards Accountable Reinforcement Learning

2018-11-07 · Christoph Dann, Lihong Li, Wei Wei, Emma Brunskill

The performance of a reinforcement learning algorithm can vary drastically during learning because of exploration. Existing algorithms provide little information about the quality of their current policy before executing it, and thus have limited use in high-stakes applications like healthcare. We address this lack of accountability by proposing that algorithms output policy certificates. These certificates bound the sub-optimality and return of the policy in the next episode, allowing humans to intervene when the certified quality is not satisfactory. We further introduce two new algorithms with certificates and present a new framework for theoretical analysis that guarantees the quality of their policies and certificates. For tabular MDPs, we show that computing certificates can even improve the sample-efficiency of optimism-based exploration. As a result, one of our algorithms is the first to achieve minimax-optimal PAC bounds up to lower-order terms, and this algorithm also matches (and in some settings slightly improves upon) existing minimax regret bounds.

📄 PDF Abstract BibTeX arXiv:1811.03056

Code (0)

등록된 구현이 없습니다.

Tasks

reinforcement-learningReinforcement LearningReinforcement Learning (RL)

Similar Papers 제목 키워드 기반

Accountable Off-Policy Evaluation via a Kernelized Bellman Statistics

2020-01-01 · ICML 2020 1 · Yihao Feng, Tongzheng Ren, Ziyang Tang, Qiang Liu

Off-policy evaluation plays an important role in modern reinforcement learning. However, most of the existing off-policy evaluation only focus on the value estimation, without providing an accountable confidence interval…

Off-policy evaluation

Learning Barrier Certificates: Towards Safe Reinforcement Learning with Zero Training-time Violations

2021-08-04 · NeurIPS 2021 12 · Yuping Luo, Tengyu Ma

Training-time safety violations have been a major concern when we deploy reinforcement learning algorithms in the real world. This paper explores the possibility of safe RL algorithms with zero training-time safety viola…

reinforcement-learningReinforcement Learning (RL)Safe Reinforcement Learning

PAC-Bayesian Reinforcement Learning Trains Generalizable Policies

2025-10-12 · Abdelkrim Zitouni, Mehdi Hennequin, Juba Agoun, Ryan Horache 외 arxiv

We derive a novel PAC-Bayesian generalization bound for reinforcement learning that explicitly accounts for Markov dependencies in the data, through the chain's mixing time. This contributes to overcoming challenges in o…

Reinforcement LearningContinuous Control

MSACL: Multi-Step Actor-Critic Learning with Lyapunov Certificates for Exponentially Stabilizing Control

2025-12-31 · Yongwei Zhang, Yuanzhe Xing, Quanyi Liang, Quan Quan 외 arxiv

For stabilizing control tasks, model-free reinforcement learning (RL) approaches face numerous challenges, particularly regarding the issues of effectiveness and efficiency in complex high-dimensional environments with l…

Reinforcement Learning

IV-Posterior: Inverse Value Estimation for Interpretable Policy Certificates

2020-11-30 · Tatiana Lopez-Guevara, Michael Burke, Nicholas K. Taylor, Kartic Subr

Model-free reinforcement learning (RL) is a powerful tool to learn a broad range of robot skills and policies. However, a lack of policy interpretability can inhibit their successful deployment in downstream applications…

Reinforcement Learning (RL)