Game informatics: Search of And-Or tree and Computational Complexity of games and puzzles

本研究涉及游戏信息学。（2）博弈和谜题的计算复杂性。计算机对与或树（博弈树）的搜索一般使用评价函数。 我们定义我们的游戏树模型，并比较的配置数量进行搜索的深度优先搜索在指定的深度，我们的游戏树下的各种评价功能。 结果表明，当采用概率为p（0<p<1）的最佳评价函数时，系统的配置数比采用最佳评价函数时的配置数多出多项式倍，得到了一些游戏和谜题的计算复杂度，相关结果已发表在论文或国际会议上.

This research is concerned with Game Informatics. The themes are (1) computer search of And-Or trees, and (2) computational complexity of games and puzzles.Search of And-Or tree (game-tree) by computers generally use evaluation function. We define our game-tree model, and compare numbers of configurations to be searched by depth-first search at the designated depth of our game-tree under various evaluation functions. Our result shows that the number of configurations when an evaluation function is applied, which is as close as the perfect evaluation function with probability p (0<p<1), is polynomial time larger than that of configurations when the perfect evaluation function is applied.We obtained computational complexities of some games and puzzles, and the results have been published as papers, or presented in international conferences.