Group-Agent Reinforcement Learning

Kaiyue Wu, Xiao-Jun Zeng

It can largely benefit the reinforcement learning (RL) process of each agent if multiple geographically distributed agents perform their separate RL tasks cooperatively. Different…