Belief Re-Use in Partially Observable Monte Carlo Tree Search

Ebert Theeuwes, Gabriele Venturato, Gavin Rens

2025

Abstract

Partially observable Markov decision processes (POMDPs) require agents to make decisions with incomplete information, facing challenges like an exponential growth in belief states and action-observation histories. Monte Carlo tree search (MCTS) is commonly used for this, but it redundantly evaluates identical states reached through different action sequences. We propose Belief Re-use in Online Partially Observable Planning (BROPOP), a technique that transforms the MCTS tree into a graph by merging nodes with similar beliefs. Using a POMDP-specific locality-sensitive hashing method, BROPOP efficiently identifies and reuses belief nodes while preserving information integrity through update-descent backpropagation. Experiments on standard benchmarks show that BROPOP enhances reward performance with controlled computational cost.

Paper Citation

in Harvard Style

Theeuwes E., Venturato G. and Rens G. (2025). Belief Re-Use in Partially Observable Monte Carlo Tree Search. In Proceedings of the 17th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART; ISBN 978-989-758-737-5, SciTePress, pages 634-645. DOI: 10.5220/0013265200003890

in Bibtex Style

@conference{icaart25,
author={Ebert Theeuwes and Gabriele Venturato and Gavin Rens},
title={Belief Re-Use in Partially Observable Monte Carlo Tree Search},
booktitle={Proceedings of the 17th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART},
year={2025},
pages={634-645},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0013265200003890},
isbn={978-989-758-737-5},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 17th International Conference on Agents and Artificial Intelligence - Volume 2: ICAART
TI - Belief Re-Use in Partially Observable Monte Carlo Tree Search
SN - 978-989-758-737-5
AU - Theeuwes E.
AU - Venturato G.
AU - Rens G.
PY - 2025
SP - 634
EP - 645
DO - 10.5220/0013265200003890
PB - SciTePress