Learning Opponent Strategies through First Order Induction

In a competitive game it is important to identify the opponent’s strategy as quickly and accurately as possible so that an effective response can be staged. In this vain, this paper summarizes our work in exploring the use of the first order inductive learning (FOIL) algorithm for learning rules which can be used to represent opponent strategies. Specifically, we use these learned rules to perform plan recognition and classify an opponent strategy as one of multiple learned strategies. Our experiments validate this novel approach in a simple realtime strategy game.

• Type: Journal
• Year: 2011
• Where: FLAIRS
• Authors: Katie Long Genter, Santiago Ontañón, Ashwin Ram
• Keywords: Artificial Intelligence, FLAIRS 2011, Learned Rules, Learned Strategies, Realtime Strategy Game