Efficient Deep Reinforcement Learning via Policy-Extended Successor Feature Approximator.
Yining LiTianpei YangJianye HaoYan ZhengHongyao TangPublished in: DAI (2022)
Keyphrases
- reinforcement learning
- optimal policy
- function approximators
- action selection
- reinforcement learning algorithms
- policy search
- multi agent
- dynamic programming
- actor critic
- image features
- cost effective
- markov decision processes
- approximate dynamic programming
- average cost
- function approximation
- machine learning
- state space
- learning algorithm