A perspective on off-policy evaluation in reinforcement learning.
Lihong LiPublished in: Frontiers Comput. Sci. (2019)
Keyphrases
- policy evaluation
- reinforcement learning
- temporal difference
- least squares
- model free
- monte carlo
- markov decision processes
- function approximation
- policy iteration
- td learning
- variance reduction
- reinforcement learning algorithms
- semi parametric
- optimal policy
- state space
- machine learning
- evaluation function
- learning algorithm
- dynamic programming
- multi agent
- neural network
- step size
- supervised learning
- transfer learning
- gaussian process
- statistical inference
- partially observable
- partially observable markov decision processes
- action space
- function approximators
- average reward
- state action
- decision making
- action selection