Functional Reward Markov Decision Processes: Theory and Applications.
Paul WengOlivier SpanjaardPublished in: Int. J. Artif. Intell. Tools (2017)
Keyphrases
- markov decision processes
- average reward
- reinforcement learning
- reward function
- state space
- total reward
- finite state
- discounted reward
- optimal policy
- expected reward
- policy iteration
- transition matrices
- dynamic programming
- decision processes
- reinforcement learning algorithms
- planning under uncertainty
- average cost
- markov decision process
- finite horizon
- factored mdps
- decision theoretic planning
- model based reinforcement learning
- semi markov decision processes
- stationary policies
- infinite horizon
- partially observable
- state and action spaces
- action sets
- reachability analysis
- policy gradient
- long run
- multistage
- action space
- sufficient conditions