Monte Carlo Tree Search in Chinese Dark Chess

• Item type: Symposium(1)
• 公開日: 2011-11-04
• タイトル: Monte Carlo Tree Search in Chinese Dark Chess
• タイトル / 言語: en
• タイトル: Monte Carlo Tree Search in Chinese Dark Chess
• 言語: eng
• 資源タイプ / 資源タイプ識別子: http://purl.org/coar/resource_type/c_5794
• 資源タイプ: conference paper
• 著者所属: Department of Computer Science & Information Engineering, National Dong Hwa University
• 著者所属: Department of Computer Science & Information Engineering, National Dong Hwa University
• 著者所属: Department of Computer Science & Information Engineering, National Dong Hwa University
• 著者所属: Department of Information Management, Chang Jung Christian University
• 著者所属: Department of Information Management, Chang Jung Christian University
• 著者所属: Department of AM, Chung Yuan Christian University
• 著者所属(英): en
• 著者所属(英): Department of Computer Science & Information Engineering, National Dong Hwa University
• 著者所属(英): en
• 著者所属(英): Department of Computer Science & Information Engineering, National Dong Hwa University
• 著者所属(英): en
• 著者所属(英): Department of Computer Science & Information Engineering, National Dong Hwa University
• 著者所属(英): en
• 著者所属(英): Department of Information Management, Chang Jung Christian University
• 著者所属(英): en
• 著者所属(英): Department of Information Management, Chang Jung Christian University
• 著者所属(英): en
• 著者所属(英): Department of AM, Chung Yuan Christian University
• 著者名: Shi-JimYen; Cheng-WeiChou; Tsan-ChengSu; Shun-ChinHsu; Hsin-HungChou; Jr-ChangChen
• 著者名(英): Shi-Jim, Yen; Cheng-Wei, Chou; Tsan-Cheng, Su; Shun-Chin, Hsu; Hsin-Hung, Chou; Jr-Chang, Chen
• 論文抄録 / 内容記述タイプ: Other
• 論文抄録 / 内容記述: In this paper we propose an efficient method to apply Monte Carlo Tree Search (MCTS) to Chinese dark chess (CDC) and get very well experimental results. CDC is an imperfect information variety of Chinese chess. This game is very popular in Chinese culture sphere. The key point of solving this game is how to handle the imperfect information part very well. Unfortunately, most of state-of-the-art CDC programs cannot do it. MCTS is a famous best-first search algorithm, which is suitable to make decisions in uncertain environments. We try to change the tree structure of MCTS to handle the imperfect information part of CDC, and obtain great improvement of CDC. Moreover, our program won every game in computer CDC tournament in TCGA 2011 computer game tournament.
• 論文抄録(英) / 内容記述タイプ: Other
• 論文抄録(英) / 内容記述: In this paper we propose an efficient method to apply Monte Carlo Tree Search (MCTS) to Chinese dark chess (CDC) and get very well experimental results. CDC is an imperfect information variety of Chinese chess. This game is very popular in Chinese culture sphere. The key point of solving this game is how to handle the imperfect information part very well. Unfortunately, most of state-of-the-art CDC programs cannot do it. MCTS is a famous best-first search algorithm, which is suitable to make decisions in uncertain environments. We try to change the tree structure of MCTS to handle the imperfect information part of CDC, and obtain great improvement of CDC. Moreover, our program won every game in computer CDC tournament in TCGA 2011 computer game tournament.
• 書誌情報: ゲームプログラミングワークショップ2011論文集; 巻 2011, 号 6, p. 41-45, 発行日 2011-10-28
• 出版者 / 言語: ja
• 出版者: 情報処理学会