Bi-Directional Monte Carlo Tree Search

This paper describes a new algorithm called Bi-Directional Monte Carlo Tree Search. The essential idea of Bi-directional Monte Carlo Tree Search is to run an MCTS forwards from the start state, and simultaneously run an MCTS backwards from the goal state, and stop when the two searches meet. Bi-Di...

Full description

Bibliographic Details
Main Author: Spoerer, Kristian
Format: Article
Language:English
Published: Penerbit Universiti Kebangsaan Malaysia 2021
Online Access:http://journalarticle.ukm.my/16841/
http://journalarticle.ukm.my/16841/1/02.pdf
_version_ 1848814156401082368
author Spoerer, Kristian
author_facet Spoerer, Kristian
author_sort Spoerer, Kristian
building UKM Institutional Repository
collection Online Access
description This paper describes a new algorithm called Bi-Directional Monte Carlo Tree Search. The essential idea of Bi-directional Monte Carlo Tree Search is to run an MCTS forwards from the start state, and simultaneously run an MCTS backwards from the goal state, and stop when the two searches meet. Bi-Directional MCTS is tested on 8-Puzzle and Pancakes Problem, two single-agent search problems, which allow control over the optimal solution length d and average branching factor b respectively. Preliminary results indicate that enhancing Monte Carlo Tree Search by making it Bi-Directional speeds up the search. The speedup of Bi-directional MCTS grows with increasing the problem size, in terms of both optimal solution length d and also branching factor b. Furthermore, Bi-Directional Search has been applied to a Reinforcement Learning algorithm. It is hoped that the speed enhancement of Bi-directional Monte Carlo Tree Search will also apply to other planning problems.
first_indexed 2025-11-15T00:29:36Z
format Article
id oai:generic.eprints.org:16841
institution Universiti Kebangasaan Malaysia
institution_category Local University
language English
last_indexed 2025-11-15T00:29:36Z
publishDate 2021
publisher Penerbit Universiti Kebangsaan Malaysia
recordtype eprints
repository_type Digital Repository
spelling oai:generic.eprints.org:168412021-06-20T04:56:23Z http://journalarticle.ukm.my/16841/ Bi-Directional Monte Carlo Tree Search Spoerer, Kristian This paper describes a new algorithm called Bi-Directional Monte Carlo Tree Search. The essential idea of Bi-directional Monte Carlo Tree Search is to run an MCTS forwards from the start state, and simultaneously run an MCTS backwards from the goal state, and stop when the two searches meet. Bi-Directional MCTS is tested on 8-Puzzle and Pancakes Problem, two single-agent search problems, which allow control over the optimal solution length d and average branching factor b respectively. Preliminary results indicate that enhancing Monte Carlo Tree Search by making it Bi-Directional speeds up the search. The speedup of Bi-directional MCTS grows with increasing the problem size, in terms of both optimal solution length d and also branching factor b. Furthermore, Bi-Directional Search has been applied to a Reinforcement Learning algorithm. It is hoped that the speed enhancement of Bi-directional Monte Carlo Tree Search will also apply to other planning problems. Penerbit Universiti Kebangsaan Malaysia 2021-06 Article PeerReviewed application/pdf en http://journalarticle.ukm.my/16841/1/02.pdf Spoerer, Kristian (2021) Bi-Directional Monte Carlo Tree Search. Asia-Pacific Journal of Information Technology and Multimedia, 10 (1). pp. 17-26. ISSN 2289-2192 https://www.ukm.my/apjitm/articles-year.php
spellingShingle Spoerer, Kristian
Bi-Directional Monte Carlo Tree Search
title Bi-Directional Monte Carlo Tree Search
title_full Bi-Directional Monte Carlo Tree Search
title_fullStr Bi-Directional Monte Carlo Tree Search
title_full_unstemmed Bi-Directional Monte Carlo Tree Search
title_short Bi-Directional Monte Carlo Tree Search
title_sort bi-directional monte carlo tree search
url http://journalarticle.ukm.my/16841/
http://journalarticle.ukm.my/16841/
http://journalarticle.ukm.my/16841/1/02.pdf