Reachability in recursive Markov decision processes.
Tomás BrázdilVáclav BrozekVojtech ForejtAntonín KuceraPublished in: Inf. Comput. (2008)
Keyphrases
- markov decision processes
- state space
- finite state
- optimal policy
- reinforcement learning
- policy iteration
- dynamic programming
- transition matrices
- partially observable
- model based reinforcement learning
- reinforcement learning algorithms
- reachability analysis
- heuristic search
- factored mdps
- decision theoretic planning
- average cost
- markov decision process
- finite horizon
- risk sensitive
- average reward
- planning under uncertainty
- infinite horizon
- dynamical systems
- decision processes
- discounted reward
- reward function
- action sets
- total reward
- action space