Information-guided Planning : An Online Approach for Partially Observable Problems

do Carmo Alves, Matheus Aparecido and Varma, Amokh and Elkhatib, Yehia and Soriano Marcolino, Leandro (2023) Information-guided Planning : An Online Approach for Partially Observable Problems. In: Thirty-seventh Conference on Neural Information Processing Systems :. UNSPECIFIED, USA. (In Press)

Text (Main paper - Author version)
alves2023ibpomcp.pdf - Accepted Version
Available under License Creative Commons Attribution.
Download (1MB)

Abstract

This paper presents IB-POMCP, a novel algorithm for online planning under partial observability. Our approach enhances the decision-making process by using estimations of the world belief's entropy to guide a tree search process and surpass the limitations of planning in scenarios with sparse reward configurations. By performing what we denominate as an information-guided planning process, the algorithm, which incorporates a novel I-UCB function, shows significant improvements in reward and reasoning time compared to state-of-the-art baselines in several benchmark scenarios, along with theoretical convergence guarantees.