Location:
Search - tree Table ja
Search list
Description: 使用java编写的GUI的黑白棋游戏，搜索算法采用经典的博弈树，并在此基础上做了大量优化，我的评估函数采用了Simon M. Lucas 和 Thomas P. Runarsson 在其合作发表的 Temporal Difference Learning Versus Co-Evolution for Acquiring Othello Position Evaluation 中通过对比即时差分学习（TDL，Temporal Difference
Learning）和协同进化（CEL，Co-Evolution）对于计算黑白棋的估值函数时得到的一个最佳 WPC，估值函数同时包括行动力的计算。于此同时，对于计算可行位置，我采用了基于查表的方法。对于开局，我使用了 Kyung-Joong Kim 和 Sung-Bae Cho 的 Evolutionary Othello Players Boosted by Opening Knowledge 使用的 99 种 well-defined 的开局表。
总体来说，AI智能很强，默认等级一般人是下不过它的。-GUI using java prepared Riversi games, search algorithms using classical game tree, and on this basis have done a lot of optimization, the evaluation function I used Simon M. Lucas and Thomas P. Runarsson published in its Temporal Difference Learning Versus Co-Evolution for Acquiring Othello Position Evaluation in the difference by comparing the real-time learning (TDL, Temporal Difference
Learning) and co-evolution (CEL, Co-Evolution) Smileys for the calculation of the valuation function to be one of the best of the WPC, the valuation function at the same time including the calculation of mobility. At the same time, a viable location for the calculation, I used the method based on the look-up table. For start, I used the Kyung-Joong Kim and Sung-Bae Cho s Evolutionary Othello Players Boosted by Opening Knowledge use of 99 kinds of well-defined start of the table.
Generally speaking, AI very smart, the default level, however most people are under it.
Platform: |
Size: 3481600 |
Author: da |
Hits: