Policy Gradient for Rectangular Robust Markov Decision Processes.
Navdeep KumarEsther DermanMatthieu GeistKfir Y. LevyShie MannorPublished in: NeurIPS (2023)
Keyphrases
- markov decision processes
- policy gradient
- reinforcement learning algorithms
- average reward
- reinforcement learning
- policy iteration
- state space
- actor critic
- finite state
- optimal policy
- dynamic programming
- infinite horizon
- partially observable markov decision processes
- action space
- partially observable
- reinforcement learning methods
- average cost
- markov decision process
- reward function
- computational complexity