Efficiently Training On-Policy Actor-Critic Networks in Robotic Deep Reinforcement Learning with Demonstration-like Sampled Exploration.
Zhaorun ChenBinhao ChenShenghan XieLiang GongChengliang LiuZhengfeng ZhangJunping ZhangPublished in: CoRR (2021)
Keyphrases
- actor critic
- reinforcement learning
- policy gradient
- temporal difference
- reinforcement learning algorithms
- approximate dynamic programming
- policy iteration
- optimal control
- neuro fuzzy
- function approximation
- action selection
- policy gradient methods
- markov decision processes
- model free
- supervised learning
- gradient method
- natural actor critic
- state space
- rl algorithms
- average reward
- optimal policy
- evaluation function
- mobile robot
- learning problems
- reinforcement learning problems
- step size
- transfer learning
- fuzzy logic
- temporal difference learning
- reinforcement learning methods
- semi supervised
- dynamic programming
- training set
- monte carlo
- learning algorithm