More Trees or Larger Trees: Parallelizing Monte Carlo Tree Search

Erik Steinmetz, Maria Gini

Research output: Contribution to journalArticlepeer-review

3 Scopus citations

Abstract

Monte Carlo tree search (MCTS) is being effectively used in many domains, but acquiring good results from building larger trees takes time that can in many cases be impractical. In this article, we show that parallelizing the tree building process using multiple independent trees (root parallelization) can improve results when limited time is available, and compare these results to other parallelization techniques and to results obtained from running for an extended time. We obtained our results using MCTS in the domain of computer Go, which has the most mature implementations. Compared to previous studies, our results are more precise and statistically significant.

Original languageEnglish (US)
Pages (from-to)315-320
Number of pages6
JournalIEEE Transactions on Games
Volume13
Issue number3
DOIs
StatePublished - Jan 1 2021

Bibliographical note

Funding Information:
Institute (MSI) for the use of their computing facilities, and D. Boley for his support through the project.

Publisher Copyright:
© 2018 IEEE.

Keywords

  • Game of Go
  • Monte Carlo tree search (MCTS)
  • parallelization

Cite this