Game Player Strategy Pattern Recognition and How UCT Algorithms Apply Pre-Knowledge of Player's Strategy to Improve Opponent AI

被引：10

作者：

He, Suoju ^{[1
]}

Wang, Yi ^{[1
]}

Xie, Fan ^{[1
]}

Meng, Jin ^{[1
]}

Chen, Hongtao ^{[1
]}

Luo, Sai ^{[1
]}

Liu, Zhiqing ^{[1
]}

Zhu, Qiliang ^{[1
]}

机构：

[1] Beijing Univ Posts & Telecommun, Sch Software Engn, Beijing 100876, Peoples R China

来源：

2008 INTERNATIONAL CONFERENCE ON COMPUTATIONAL INTELLIGENCE FOR MODELLING CONTROL & AUTOMATION, VOLS 1 AND 2 | 2008年

关键词：

Player Strategy; Pattern Recognition; KNN; knowledge-based UCT; Adaptive Game AI; Dead End;

D O I：

10.1109/CIMCA.2008.82

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

Player Strategy Pattern Recognition (PSPR) is to apply pattern recognition and its approach to identification of player's strategy during the gameplay. Correctly identified player's strategy, which is called knowledge, could be used to improve game opponent AI which can be implemented by KB-UCT (knowledge-based Upper Confidence bound for Trees). KB-UCT improves adaptability of game AI, the challenge level of the gameplay, and the performance of the opponent AI; as a result the entertainment of game is promoted. In this paper, the prey and predator game genre of Dead End game is used as a test-bed. During the PSPR, classification algorithm of KNN (k-nearest neighbor) is chosen to analyze off-line data from the simulated gamers who are choosing different strategies. Based on the information from PSPR, the game AI is promoted through application of KB-UCT, in this case, domain knowledge is used for UCT tree pruning; as a result the performance of the opponent AI is enhanced.

引用

页码：1177 / 1181

页数：5