モンテカルロ木探索を用いたぷよぷよAI

添島 克駿; 山口 暢彦; 福田 修; 奥村 浩

doi:10.14864/fss.36.0_527

36th Fuzzy System Symposium

Session ID : WD2-1

DOI https://doi.org/10.14864/fss.36.0_527

Conference information

Host: Japan Society for Fuzzy Theory and Intelligent Informatics (SOFT)

Name : 36th Fuzzy System Symposium

Number : 36

Location : [in Japanese]

Date : September 07, 2020 - September 09, 2020

proceeding

Puyo Puyo AI using Monte Carlo tree search

*Katsutoshi Soejima, Nobuhiko Yamaguchi, Osamu Fukuda, Hiroshi Okumura

Author information

CONFERENCE PROCEEDINGS FREE ACCESS

Details

Abstract

In this paper, we develop the incomplete information game Puyo Puyo AI. We introduce Monte Carlo tree search to Puyo Puyo AI instead of exhaused search with a board evaluation system. Our goal is to improve the search speed by giving priority to the high scoring boards.

Corresponding author

Conference information

Register with J-STAGE for free!