Strategy Classification in Multi-agent Environment - Applying Reinforcement Learning to Soccer Agents -
Eiji Uchibe, Minoru Asada, Koh Hosoda · 1996
This paper proposes a method for agent behavior clas-sification which estimates the relations between the learner’s behaviors and the other agents in the en-vironment through interactions using the method of system identification. In order to identify the mod-el of each agent, Akaike’s Information Criterion(AIC) is applied to the result of Canonical Variate Analy-sis(CVA). Next, reinforcement learning based on the estimated state vectors is used in order to obtain the optimal behavior. The proposed method is applied to soccer playing robots. Unlike our previous work, the method can cope with a rolling ball. Computer sim-ulations and preliminary experiments are shown and the discussion is given.