研究生: |
張子樂 Tzu-Le Chang |
---|---|
論文名稱: |
對手行為預測於麻將之應用 Application of Opponent Behavior Prediction in Mahjong |
指導教授: |
戴文凱
Wen-Kai Tai |
口試委員: |
姚智原
江佩穎 |
學位類別: |
碩士 Master |
系所名稱: |
電資學院 - 資訊工程系 Department of Computer Science and Information Engineering |
論文出版年: | 2020 |
畢業學年度: | 108 |
語文別: | 英文 |
論文頁數: | 48 |
中文關鍵詞: | 多人不完全信息遊戲 、預測 、rule-based策略 、麻將 |
外文關鍵詞: | multi-player games with imperfect information, prediction, rule-based strategies, Mahjong |
相關次數: | 點閱:223 下載:18 |
分享至: |
查詢本校圖書館目錄 查詢臺灣博碩士論文知識加值系統 勘誤回報 |
台灣麻將(台麻)是一個盛行於台灣的多人不完全信息博弈遊戲。在本論文中,我們提出了一種rule-based的台麻AI,它使用對手行為預測模組來優化AI。此模組根據由啟發法的台麻原理推導出且經過遊戲紀錄驗證的麻將規則,能夠預測對手需要某張麻將牌的機率。即使是在使用預測模組之前,AI已勝過平均人類玩家,胡牌率為25.455%對上21.638%。在使用此模組後,AI的表現相較於之前有大幅的提升,胡牌率為28.830%對上17.953%。我們提出的AI是容易調整、能夠減少玩家的等待時間(由於台麻需要四個人)並提升玩家的遊戲體驗。因此它很適合做為商業使用,我們的台麻AI目前正於業界所使用中。
Taiwanese Mahjong (Tai-Mah) is a popular multi-player tile-based game with imperfect information in Taiwan. In this paper we present a rule-based Tai-Mah AI using opponent behavior prediction agent to optimize the AI. The agent can predict the probability of Mahjong tiles that an opponent needs based on heuristic Tai-Mah principles and then verified by game records. The AI plays better than average human players even before using the prediction agent, with the winning percentage 25.455% over 21.638%. After implementing the agent, the performance of AI has greatly improved compared to the previous version, with the winning percentage 28.830% over 17.953%. The proposed AI is easy to adjust, it can reduce players' waiting time (since a Mahjong game requires four players) and improve players' game experience. Therefore it is suitable for commercial use, our AI is currently being used in the industry.
[1] N. Mizukami and Y. Tsuruoka, “Building a computer mahjong player based on monte carlo simulation and opponent models,” in 2015 IEEE Conference on Computational Intelligence and Games (CIG), pp. 275–283, IEEE, 2015.
[2] R. Fathzadeh, V. Mokhtari, M. Mousakhani, and F. Mahmoudi, “Mining opponent behavior: a champion of robocup coach competition,” in 2006 IEEE 3rd Latin American Robotics Symposium, pp. 80–83, IEEE, 2006.
[3] D. Churchill, M. Buro, and R. Kelly, “Robust continuous build-order optimization in starcraft,” in 2019 IEEE Conference on Games (CoG), pp. 1–8, IEEE, 2019.
[4] L. F. Teófilo, L. P. Reis, and H. L. Cardoso, “A profitable online no-limit poker playing agent,” in 2014 IEEE/WIC/ACM International Joint Conferences on Web Intelligence (WI) and Intelligent Agent Technologies (IAT), vol. 3, pp. 286–293, IEEE, 2014.
[5] L. K. Chuang and I. C. Wu, “A study of mahjong program design,” Master’s thesis, National Chiao Tung University, 2015.
[6] K. Yoshimura, T. Hochin, and H. Nomiya, “Searching optimal movements in multi-player games with imperfect information,” in 2016 IEEE/ACIS 15th International Conference on Computer and Information Science (ICIS), pp. 1–6, IEEE, 2016.
[7] K. Yoshimura, T. Hochin, and H. Nomiya, “Estimation of rates arriving at the winning hands in multi-player games with imperfect information,” in 2016 4th Intl Conf on Applied Computing and Information Technology/3rd Intl Conf on Computational Science/Intelligence and Applied Informatics/1st Intl Conf on Big Data, Cloud Computing, Data Science & Engineering (ACITCSIIBCD), pp. 99–104, IEEE, 2016.
[8] K. H. Feng and W. K. Tai, “Optimized monte carlo tree search for mahjong agent,” Master’s thesis, National Taiwan University of Science and Technology, 2019.