Large Scale Markov Decision Processes with Changing Rewards.
Adrian Rivera CardosoHe WangHuan XuPublished in: CoRR (2019)
Keyphrases
- state and action spaces
- markov decision processes
- action space
- reinforcement learning
- state space
- finite state
- transition matrices
- average reward
- discounted reward
- optimal policy
- policy iteration
- dynamic programming
- decision theoretic planning
- planning under uncertainty
- reinforcement learning algorithms
- partially observable
- factored mdps
- finite horizon
- reachability analysis
- reward function
- sequential decision making under uncertainty
- average cost
- infinite horizon
- markov decision process
- decision processes
- learning algorithm
- state abstraction
- machine learning
- optimal solution