Reward Space Noise for Exploration in Deep Reinforcement Learning.
Chuxiong SunRui WangQian LiXiaohui HuPublished in: Int. J. Pattern Recognit. Artif. Intell. (2021)
Keyphrases
- reinforcement learning
- action space
- exploration strategy
- state space
- function approximation
- action selection
- random noise
- low dimensional
- learning process
- optimal policy
- space time
- signal to noise ratio
- noise level
- active exploration
- noise reduction
- learning algorithm
- exploration exploitation
- autonomous learning
- partially observable environments
- multi agent
- policy gradient
- average reward
- multi agent systems
- search space
- additive noise
- model free
- markov decision processes
- image restoration