Research on Multi-agent Team Formation Based on Reinforcement Learning

The research actuality of reinforcement learning in inland and oversea is introduced in this paper. The method in which Q-learning and neural network are used to implement adaptive team formation of multi-agent system is presented in detail.