期刊名称：International Journal of Artificial Intelligence and Expert Systems (IJAE)
电子版ISSN：2180-124X
出版年度：2011
卷号：2
期号：1
页码：1-13
出版社：Computer Science Journals
摘要：As an example of multi-agent learning in soccer games of the RoboCup 2D Soccer Simulation League, we dealt with a learning problem between a kicker and a receiver when a direct free kick is awarded just outside the opponent's penalty area. We propose how to use a heuristic function to evaluate an advantageous target point for safely sending/receiving a pass and scoring. The heuristics include an interaction term between a kicker and a receiver to intensify their coordination. To calculate the interaction term, we let a kicker/receiver agent have a receiver's/kicker's action decision model to predict a receiver's/kicker's action. Parameters in the heuristic function can be learned by a kind of reinforcement learning called the policy gradient method. Our experiments show that if the two agents do not have the same type of heuristics, the interaction term based on prediction of a teammate's decision model leads to learning a master-servant relation between a kicker and a receiver, where a receiver is a master and a kicker is a servant.
