Computable approximations for average Markov decision processes in continuous time.
Jonatha AnselmiFrançois DufourTomás Prieto-RumeauPublished in: J. Appl. Probab. (2018)
Keyphrases
- markov decision processes
- state space
- average cost
- stationary policies
- discounted reward
- finite state
- optimal policy
- reinforcement learning
- dynamic programming
- policy evaluation
- transition matrices
- planning under uncertainty
- optimal control
- markov decision process
- markov chain
- factored mdps
- policy iteration
- infinite horizon
- finite horizon
- decision theoretic planning
- decision processes
- reinforcement learning algorithms
- dynamical systems
- average reward
- action space
- partially observable
- search space
- state abstraction
- reachability analysis
- model based reinforcement learning
- risk sensitive
- efficient computation
- state and action spaces
- semi markov decision processes
- machine learning