Saved in:
Bibliographic Details
Main Authors: Santos, Pedro P., Silvestrin, Jacopo, Sardinha, Alberto, Melo, Francisco S.
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2601.17667
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866918324210761728
author Santos, Pedro P.
Silvestrin, Jacopo
Sardinha, Alberto
Melo, Francisco S.
author_facet Santos, Pedro P.
Silvestrin, Jacopo
Sardinha, Alberto
Melo, Francisco S.
contents We propose a provably correct Monte Carlo tree search (MCTS) algorithm for solving risk-aware Markov decision processes (MDPs) with entropic risk measure (ERM) objectives. We provide a non-asymptotic analysis of our proposed algorithm, showing that the algorithm: (i) is correct in the sense that the empirical ERM obtained at the root node converges to the optimal ERM; and (ii) enjoys polynomial regret concentration. Our algorithm successfully exploits the dynamic programming formulations for solving risk-aware MDPs with ERM objectives introduced by previous works in the context of an upper confidence bound-based tree search algorithm. Finally, we provide a set of illustrative experiments comparing our risk-aware MCTS method against relevant baselines.
format Preprint
id arxiv_https___arxiv_org_abs_2601_17667
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Entropic Risk-Aware Monte Carlo Tree Search
Santos, Pedro P.
Silvestrin, Jacopo
Sardinha, Alberto
Melo, Francisco S.
Machine Learning
We propose a provably correct Monte Carlo tree search (MCTS) algorithm for solving risk-aware Markov decision processes (MDPs) with entropic risk measure (ERM) objectives. We provide a non-asymptotic analysis of our proposed algorithm, showing that the algorithm: (i) is correct in the sense that the empirical ERM obtained at the root node converges to the optimal ERM; and (ii) enjoys polynomial regret concentration. Our algorithm successfully exploits the dynamic programming formulations for solving risk-aware MDPs with ERM objectives introduced by previous works in the context of an upper confidence bound-based tree search algorithm. Finally, we provide a set of illustrative experiments comparing our risk-aware MCTS method against relevant baselines.
title Entropic Risk-Aware Monte Carlo Tree Search
topic Machine Learning
url https://arxiv.org/abs/2601.17667