Mutual Information as Intrinsic Reward of Reinforcement Learning Agents for On-demand Ride Pooling.
Xianjie ZhangJiahao SunChen GongKai WangYifei CaoHao ChenHao ChenYu LiuPublished in: CoRR (2023)
Keyphrases
- mutual information
- reinforcement learning agents
- reinforcement learning
- dynamic environments
- information theoretic
- image registration
- state abstraction
- similarity measure
- feature selection
- multi agent
- state space
- transfer learning
- multi agent environments
- reward function
- function approximation
- markov decision processes
- long run
- machine learning
- learning algorithm
- model free
- temporal difference
- average reward
- high dimensional
- cooperative