Efficient Policy Iteration for Robust Markov Decision Processes via Regularization.
Navdeep KumarKfir LevyKaixin WangShie MannorPublished in: CoRR (2022)
Keyphrases
- markov decision processes
- policy iteration
- optimal policy
- finite state
- transition matrices
- average reward
- state space
- reinforcement learning
- least squares
- infinite horizon
- model free
- factored mdps
- sample path
- dynamic programming
- policy evaluation
- average cost
- decision processes
- reinforcement learning algorithms
- planning under uncertainty
- fixed point
- finite horizon
- partially observable
- stochastic games
- markov decision process
- state and action spaces
- linear programming
- actor critic
- hierarchical reinforcement learning
- discounted reward
- markov games