| 一种围棋定式的机器学习方法 |
| |
| 引用本文： | 谷蓉,刘学民,朱仲涛,周杰.一种围棋定式的机器学习方法[J].计算机工程,2004,30(6):142-144,173. |
| |
| 作者姓名： | 谷蓉 刘学民 朱仲涛 周杰 |
| |
| 作者单位： | 1. 智能技术与系统国家重点实验室智能信息处理分室,清华大学自动化系,北京,100084
2. 智能技术与系统国家重点实验室,清华大学计算机科学与技术系,北京,100084 |
| |
| 摘 要： | 提出了一种围棋定式的机器学习方法。利用此方法可实现从棋谱库中自动提取定式并生成定式库。此外，对于棋谱数量较大的情况，采用分阶段学习方法，提高了学习效率。应用此方法，时34000局棋谱进行处理，得到定式点680638个。最后，还给出了1种基于组合博弈理论在计算机围棋博弈系统中使用定式的方法。
|
| 关 键 词： | 围棋定式 机器学习 组合博弈理论 |
| 文章编号： | 1000-3428(2004)06-0142-03 |
A Machine Learning Method of Joseki Database for Computer Go |
| |
| GU Rong,LIU Xuemin,ZHU Zhongtao,ZHOU Jie.A Machine Learning Method of Joseki Database for Computer Go[J].Computer Engineering,2004,30(6):142-144,173. |
| |
| Authors: | GU Rong LIU Xuemin ZHU Zhongtao ZHOU Jie |
| |
| Affiliation: | GU Rong1,LIU Xuemin2,ZHU Zhongtao2,ZHOU Jie1 |
| |
| Abstract: | A machine learning method of Go joseki for computer Go game system is proposed in this paper. By using this method, various Go joseki can be automatically extracted from human-played Go game files so as to construct a joseki-tree database. The learning procedure can be separated by several steps when the learning data set is too large. This approach improves the learning efficiency. 34 000 human-played Go games are processed and 680 638 joseki nodes are created. In addition, a method of how to apply joseki, which is based on combinatory game theory, to computer Go game system is presented. |
| |
| Keywords: | Joseki Machine learning Combinatorial game theory |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
