Off-Policy Evaluation for Large Action Spaces via Policy Convolution.
Noveen SachdevaLequn WangDawen LiangNathan KallusJulian J. McAuleyPublished in: CoRR (2023)
Keyphrases
- policy evaluation
- action space
- markov decision processes
- reinforcement learning
- policy iteration
- state space
- optimal policy
- markov decision problems
- temporal difference
- action selection
- least squares
- markov decision process
- dynamic programming
- finite state
- model free
- state action
- reinforcement learning algorithms
- partially observable
- average cost
- partially observable markov decision processes
- function approximation
- average reward
- real valued
- reinforcement learning methods
- reward function
- variance reduction
- semi parametric
- function approximators
- decision processes
- monte carlo
- linear programming
- policy gradient
- multi agent