Learning System Using Hierarchical Fuzzy ART for Two-Player Games

The adaptive resonance theory (ART) can generate and grow the recognition categories based on the similarity between inputs and memories. This paper proposes a new learning system using hierarchical fuzzy ART for two-player games. The proposed system segments an input state space into subspaces by the fuzzy ARTs added hierarchically in proportion as the learning progress of players, and then learns pairs of input states and actions by the reinforcement learning. As the results of experiments, it is shown through a fighting simulation game that the player can acquire proper pairs of the input states and actions against the opponent player by learning using the proposed system.