Non-Stationary Markov Decision Processes, a Worst-Case Approach using Model-Based Reinforcement Learning.
Erwan LecarpentierEmmanuel RachelsonPublished in: NeurIPS (2019)
Keyphrases
- model based reinforcement learning
- non stationary
- markov decision processes
- finite state
- state space
- reinforcement learning
- policy iteration
- optimal policy
- lower bound
- average cost
- planning under uncertainty
- dynamic programming
- partially observable
- decision processes
- np hard
- reinforcement learning algorithms
- markov decision process
- decision problems
- machine learning
- finite horizon
- infinite horizon