Finite approximation for finite-horizon continuous-time Markov decision processes.
Qingda WeiPublished in: 4OR (2017)
Keyphrases
- markov decision processes
- finite horizon
- stationary policies
- state space
- state and action spaces
- optimal policy
- optimal stopping
- markov decision process
- infinite horizon
- average cost
- finite state
- policy iteration
- dynamic programming
- reinforcement learning
- transition matrices
- action space
- markov chain
- decision theoretic planning
- partially observable
- optimal control
- average reward
- finite number
- control policies
- reward function
- dynamical systems
- lost sales
- queueing networks
- decision problems
- linear programming
- supply chain
- machine learning