Blackwell Online Learning for Markov Decision Processes.
Tao LiGuanze PengQuanyan ZhuPublished in: CISS (2021)
Keyphrases
- markov decision processes
- online learning
- state space
- finite state
- reinforcement learning
- dynamic programming
- optimal policy
- policy iteration
- transition matrices
- e learning
- average cost
- decision theoretic planning
- reinforcement learning algorithms
- partially observable
- markov decision process
- finite horizon
- decision processes
- average reward
- active learning
- factored mdps
- action space
- reachability analysis
- stochastic shortest path
- infinite horizon
- risk sensitive
- planning under uncertainty
- state and action spaces
- decision diagrams
- model based reinforcement learning
- linear programming