Finding the K best policies in a finite-horizon Markov decision process.
Lars Relund NielsenAnders Ringgaard KristensenPublished in: Eur. J. Oper. Res. (2006)
Keyphrases
- markov decision process
- finite horizon
- optimal policy
- state space
- infinite horizon
- markov decision processes
- optimal stopping
- reinforcement learning
- inventory models
- inventory control
- stochastic inventory control
- single product
- initial state
- control policies
- periodic review
- average cost
- long run
- decision problems
- finite state
- dynamic programming
- machine learning
- reward function
- optimal control
- lost sales
- multistage
- partially observable
- search algorithm