Provably Efficient Partially Observable Risk-Sensitive Reinforcement Learning with Hindsight Observation

arXiv:2402.18149v1 Announce Type: new
Abstract: This work pioneers regret analysis of risk-sensitive reinforcement learning in partially observable environments with hindsight observation, addressing a gap in theoretical exploration. We introduce a novel formulation that integrates hindsight observations into a Partially Observable Markov Decision Process (POMDP) framework, where the goal is to optimize accumulated reward under the entropic risk measure. We develop the first provably efficient RL algorithm tailored for this setting. We also prove by rigorous analysis that our algorithm achieves …

abstract analysis arxiv cs.lg decision environments exploration framework gap markov novel observable observation process reinforcement reinforcement learning risk stat.ml type work