Exploration via Distributional Reinforcement Learning with Epistemic and Aleatoric Uncertainty Estimation.
Qi LiuYanjie LiYuecheng LiuMeiling ChenShaohua LvYunhong XuPublished in: CASE (2021)
Keyphrases
- reinforcement learning
- active exploration
- action selection
- exploration strategy
- function approximation
- partial observability
- co occurrence
- uncertain data
- estimation accuracy
- estimation algorithm
- robust estimation
- model based reinforcement learning
- markov decision processes
- exploration exploitation
- state space
- neural network
- estimation error
- conceptual change
- learning tasks
- measurement error
- inherent uncertainty
- decision making