Distribution-based objectives for Markov Decision Processes.
S. AkshayBlaise GenestNikhil VyasPublished in: CoRR (2018)
Keyphrases
- markov decision processes
- state space
- finite state
- optimal policy
- reinforcement learning
- policy iteration
- transition matrices
- dynamic programming
- reinforcement learning algorithms
- partially observable
- risk sensitive
- decision theoretic planning
- infinite horizon
- factored mdps
- planning under uncertainty
- average reward
- state and action spaces
- model based reinforcement learning
- markov decision process
- probability distribution
- reachability analysis
- decision processes
- action space
- average cost
- action sets
- objective function