Existence of average optimal policies in Markov control processes with strictly unbounded costs.
Onésimo Hernández-LermaPublished in: Kybernetika (1993)
Keyphrases
- optimal policy
- average cost
- optimal control
- markov decision processes
- control policy
- production processes
- long run
- finite state
- state space
- control policies
- decision problems
- finite horizon
- markov chain
- infinite horizon
- stationary policies
- total cost
- reinforcement learning
- finite number
- multistage
- control system
- dynamic programming
- expected cost
- inventory models
- state dependent
- serial inventory systems
- discounted reward
- average reward
- initial state
- control strategy
- fixed cost
- linear programming
- machine learning
- holding cost
- inventory control
- markov decision problems
- search algorithm
- average reward reinforcement learning
- lot sizing