Monotone optimal control for a class of Markov decision processes.
Weifen ZhuangMichael Z. F. LiPublished in: Eur. J. Oper. Res. (2012)
Keyphrases
- optimal control
- markov decision processes
- risk sensitive
- dynamic programming
- infinite horizon
- average cost
- reinforcement learning
- policy iteration
- control problems
- state space
- optimal policy
- finite state
- finite horizon
- transition matrices
- decision theoretic planning
- action space
- control strategy
- policy iteration algorithm
- optimal control problems
- partially observable
- markov decision process
- average reward
- multistage
- stochastic games
- markov decision problems
- production planning
- utility function
- initial state
- reinforcement learning algorithms
- reward function