Imitation learning for combat system in RTS games with application to starcraft

Unlike the situation with regard to board games, artificial intelligence (AI) for real-time strategy (RTS) games usually suffers from an infinite number of possible future states. Furthermore, it must handle the complexity quickly. This constraint makes it difficult to build AI for RTS games with current state-of-the-art intelligent techniques. This paper proposes the use of imitation learning based on a human player's replays, which allows the AI to mimic the behaviors. During game play, the AI exploits the replay repository to search for the best similar moment from an influence map representation. This work focuses on combat in RTS games, considering the spatial configuration and unit types. Experimental results show that the proposed AI can defeat well-known competition entries a large percentage of the time.