A Reward Optimization Method Based on Action Subrewards in Hierarchical Reinforcement Learning

Figure 3

The performance comparison between divide and rule policy and nondivide and rule policy.