Provably Efficient Long-Horizon Exploration in Monte Carlo Tree Search through State Occupancy Regularization

Liam Schramm, Abdeslam Boularias
Proceedings of the 41st International Conference on Machine Learning, PMLR 235:43828-43861, 2024.

Abstract

Monte Carlo tree search (MCTS) has been successful in a variety of domains, but faces challenges with long-horizon exploration when compared to sampling-based motion planning algorithms like Rapidly-Exploring Random Trees. To address these limitations of MCTS, we derive a tree search algorithm based on policy optimization with state-occupancy measure regularization, which we call Volume-MCTS. We show that count-based exploration and sampling-based motion planning can be derived as approximate solutions to this state-occupancy measure regularized objective. We test our method on several robot navigation problems, and find that Volume-MCTS outperforms AlphaZero and displays significantly better long-horizon exploration properties.

Cite this Paper

BibTeX
@InProceedings{pmlr-v235-schramm24a, title = {Provably Efficient Long-Horizon Exploration in {M}onte {C}arlo Tree Search through State Occupancy Regularization}, author = {Schramm, Liam and Boularias, Abdeslam}, booktitle = {Proceedings of the 41st International Conference on Machine Learning}, pages = {43828--43861}, year = {2024}, editor = {Salakhutdinov, Ruslan and Kolter, Zico and Heller, Katherine and Weller, Adrian and Oliver, Nuria and Scarlett, Jonathan and Berkenkamp, Felix}, volume = {235}, series = {Proceedings of Machine Learning Research}, month = {21--27 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v235/main/assets/schramm24a/schramm24a.pdf}, url = {https://proceedings.mlr.press/v235/schramm24a.html}, abstract = {Monte Carlo tree search (MCTS) has been successful in a variety of domains, but faces challenges with long-horizon exploration when compared to sampling-based motion planning algorithms like Rapidly-Exploring Random Trees. To address these limitations of MCTS, we derive a tree search algorithm based on policy optimization with state-occupancy measure regularization, which we call Volume-MCTS. We show that count-based exploration and sampling-based motion planning can be derived as approximate solutions to this state-occupancy measure regularized objective. We test our method on several robot navigation problems, and find that Volume-MCTS outperforms AlphaZero and displays significantly better long-horizon exploration properties.} }
Endnote
%0 Conference Paper %T Provably Efficient Long-Horizon Exploration in Monte Carlo Tree Search through State Occupancy Regularization %A Liam Schramm %A Abdeslam Boularias %B Proceedings of the 41st International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2024 %E Ruslan Salakhutdinov %E Zico Kolter %E Katherine Heller %E Adrian Weller %E Nuria Oliver %E Jonathan Scarlett %E Felix Berkenkamp %F pmlr-v235-schramm24a %I PMLR %P 43828--43861 %U https://proceedings.mlr.press/v235/schramm24a.html %V 235 %X Monte Carlo tree search (MCTS) has been successful in a variety of domains, but faces challenges with long-horizon exploration when compared to sampling-based motion planning algorithms like Rapidly-Exploring Random Trees. To address these limitations of MCTS, we derive a tree search algorithm based on policy optimization with state-occupancy measure regularization, which we call Volume-MCTS. We show that count-based exploration and sampling-based motion planning can be derived as approximate solutions to this state-occupancy measure regularized objective. We test our method on several robot navigation problems, and find that Volume-MCTS outperforms AlphaZero and displays significantly better long-horizon exploration properties.
APA
Schramm, L. & Boularias, A.. (2024). Provably Efficient Long-Horizon Exploration in Monte Carlo Tree Search through State Occupancy Regularization. Proceedings of the 41st International Conference on Machine Learning, in Proceedings of Machine Learning Research 235:43828-43861 Available from https://proceedings.mlr.press/v235/schramm24a.html.

Related Material