Monge Properties, Optimal Greedy Policies, and Policy Improvement for the Dynamic Stochastic Transportation Problem.
Alexander S. EstesMichael O. BallPublished in: INFORMS J. Comput. (2021)
Keyphrases
- transportation problem
- control policies
- optimal policy
- echelon stock
- dynamic programming
- control policy
- locally optimal
- state dependent
- allocation policy
- fixed charge
- reinforcement learning
- scheduling policies
- infinite horizon
- average cost
- markov decision process
- finite horizon
- asymptotically optimal
- inventory policy
- optimal solution
- holding cost
- periodic review
- management policies
- action space
- access control policies
- base stock policies
- search algorithm
- pricing strategies
- revenue management
- multistage
- expected reward
- inventory replenishment
- expected cost