Optimality of Mixed Policies for Average Continuous-Time Markov Decision Processes with Constraints.
Xianping GuoYi ZhangPublished in: Math. Oper. Res. (2016)
Keyphrases
- markov decision processes
- stationary policies
- average cost
- optimal policy
- state space
- finite state
- action sets
- markov decision process
- optimal control
- average reward
- policy iteration
- dynamic programming
- infinite horizon
- discounted reward
- finite horizon
- reinforcement learning algorithms
- lot sizing
- reinforcement learning
- long run
- linear program
- transition matrices
- reachability analysis
- initial state
- total reward
- risk sensitive
- reward function
- policy iteration algorithm
- partially observable
- model based reinforcement learning
- markov chain
- factored mdps
- finite number
- decision processes
- decentralized control
- multistage
- state abstraction
- decision theoretic planning
- decision problems
- sufficient conditions
- state and action spaces
- search space
- machine learning
- total cost
- control policies
- planning under uncertainty