Adaptive Fighting Game Computer Player by Switching Multiple Rule-based Controllers

This paper proposes the design of a computer player for fighting games that has the advantages of both rule-based and online machine learning players. This method combines multiple computer players and switches control of the fighting character among them at regular time intervals. In this way the computer player as a whole tries to act advantageously against the current opponent player. To select appropriate controllers against the opponent out of the multiple players, we use the Sliding Window Upper Confidence Bound (SW-UCB) algorithm that is designed for non-stationary multi-armed bandit problems. We use the FightingICE platform as a testbed for our proposed method. Some experiments show the effectiveness of our proposed method in fighting games. The computer player consists of three rule-based computer players, and our method outperforms each of the three players. Additionally the proposed method improves the performance slightly against an online machine learning player.