A Dynamic Adjusting Reward Function Method for Deep Reinforcement Learning with Adjustable Parameters

Figure 3

Success rate of DQN and DQN with the idea.