摘要：Most of the artificial intelligence implementations in fighting games is based on procedural artificial intelligence, which makes the system only act based on the rules that already configured. System would not have the capacity to adapt with the user’s playing patterns. By using monotonic selection, one of the methods of fuzzy inference engine, connection between position and attack type variables with user’s playing pattern are shown. The result is, when the system are tested by playing with user’s character with a different playing pattern, the system playing pattern’s changed according to the rules that were decided through the knowledge acquisition.
