Temporal-difference emphasis learning with regularized correction for off-policy evaluation and control.
Jiaqing CaoQuan LiuLan WuQiming FuShan ZhongPublished in: Appl. Intell. (2023)
Keyphrases
- temporal difference
- td learning
- reinforcement learning
- policy evaluation
- action selection
- function approximation
- least squares
- evaluation function
- model free
- monte carlo
- learning algorithm
- supervised learning
- learning process
- policy iteration
- active learning
- step size
- statistical learning
- learning tasks
- markov decision processes
- convergence speed
- function approximators