Provably Safe Reinforcement Learning for Stochastic Reach-Avoid Problems with Entropy Regularization

2026-01-13 · Abhijit Mazumdar, Rafal Wisniewski, Manuela L. Bujorianu arxiv

We consider the problem of learning the optimal policy for Markov decision processes with safety constraints. We formulate the problem in a reach-avoid setup. Our goal is to design online reinforcement learning algorithms that ensure safety constraints with arbitrarily high probability during the learning phase. To this end, we first propose an algorithm based on the optimism in the face of uncertainty (OFU) principle. Based on the first algorithm, we propose our main algorithm, which utilizes entropy regularization. We investigate the finite-sample analysis of both algorithms and derive their regret bounds. We demonstrate that the inclusion of entropy regularization improves the regret and drastically controls the episode-to-episode variability that is inherent in OFU-based safe RL algorithms.

📄 PDF Abstract BibTeX arXiv:2601.08646

Code (0)

등록된 구현이 없습니다.

Tasks

Reinforcement Learning

Stochastic Minimum-Cost Reach-Avoid Reinforcement Learning

2026-05-12 · Jingduo Pan, Taoran Wu, Yiling Xue, Bai Xue arxiv

We study stochastic minimum-cost reach-avoid reinforcement learning, where an agent must satisfy a reach-avoid specification with probability at least $p$ while minimizing expected cumulative costs in stochastic environm…

Reinforcement Learning

Provably Safe PAC-MDP Exploration Using Analogies

2020-07-07 · Melrose Roderick, Vaishnavh Nagarajan, J. Zico Kolter

A key challenge in applying reinforcement learning to safety-critical domains is understanding how to balance exploration (needed to attain good performance on the task) with safety (needed to avoid catastrophic failure)…

reinforcement-learning Reinforcement Learning Reinforcement Learning (RL)Safe Exploration

Learning Control Policies for Stochastic Systems with Reach-avoid Guarantees

2022-10-11 · Đorđe Žikelić, Mathias Lechner, Thomas A. Henzinger, Krishnendu Chatterjee

We study the problem of learning controllers for discrete-time non-linear stochastic dynamical systems with formal reach-avoid guarantees. This work presents the first method for providing formal reach-avoid guarantees, …

Provably Safe Reinforcement Learning via Action Projection using Reachability Analysis and Polynomial Zonotopes

2022-10-19 · Niklas Kochdumper, Hanna Krasowski, Xiao Wang, Stanley Bak 외

While reinforcement learning produces very promising results for many applications, its main disadvantage is the lack of safety guarantees, which prevents its use in safety-critical systems. In this work, we address this…

reinforcement-learning Reinforcement Learning Reinforcement Learning (RL)Safe Reinforcement Learning

Provably Safe Finite-Time Guidance for Marine Vehicles

2024-02-09 · Bhawana Singh, Karim Ahmadi Dastgerdi, Nikolaos Athanasopoulos, Wasif Naeem 외

We consider a new control strategy for marine navigation, equipped with finite-time convergence characteristics. We provide mathematical guarantees for waypoint reaching and obstacle avoidance for different encounter sce…