Constrained Markov Decision Processes with Total Expected Cost Criteria.
Eitan AltmanSaid BoularoukDidier JosselinPublished in: VALUETOOLS (2019)
Keyphrases
- markov decision processes
- expected cost
- long run
- optimal policy
- cost function
- reinforcement learning
- state space
- transition matrices
- finite state
- policy iteration
- optimal strategy
- decision theoretic planning
- finite horizon
- average cost
- dynamic programming
- infinite horizon
- holding cost
- inventory level
- special case
- markov decision process
- average reward
- objective function
- sufficient conditions
- state dependent
- reward function
- finite number
- supply chain