A new multi-agent reinforcement learning approach

A new multi-agent reinforcement learning approach is proposed to learn the optimal behaviors among cooperative agent teams. The approach combines advantages of the integer programming, single agent learning and repeated game in a multi-agent framework. The integer programming is used to build cooperative teams in order to prevent the curse of dimensionality. Every cooperative team learns independently, whose members take the best response actions in the light of other agents actions in the same condition, after many repeated games, the aim root could be found. Because of other agents influence, the process of learning is supervised periodically, then through changing the learning rate to gain the right learning results. Simulation results on pursuit problem show that the proposed learning approach overcomes the divergence and improves learning speed obviously.