Continuous-Time Markov Decision Processes with Discounted Rewards: The Case of Polish Spaces.
Xianping GuoPublished in: Math. Oper. Res. (2007)
Keyphrases
- markov decision processes
- state space
- finite state
- optimal policy
- reinforcement learning
- dynamic programming
- average cost
- infinite horizon
- policy iteration
- average reward
- decision processes
- planning under uncertainty
- stationary policies
- transition matrices
- reward function
- finite horizon
- decision theoretic planning
- sequential decision making under uncertainty
- markov decision process
- reinforcement learning algorithms
- risk sensitive
- model based reinforcement learning
- factored mdps
- discounted reward
- partially observable
- optimal control
- dynamical systems
- markov chain
- reachability analysis
- state and action spaces
- action sets
- machine learning
- multi agent
- search space
- decentralized control
- state abstraction
- least squares