Hierarchical Multi-Agent Reinforcement Learning with Extensible Cooperation Graph

Proposing a novel Hierarchical Cooperation Graph Learning (HCGL) model to address complex multi-agent problems with hierarchical cooperation.