Conclusions and Open Research Problems
The Markovian demand approach provides a realistic way of modeling real-world demand scenarios. It allows us to relax the common assumption of demands being independent over time in the inventory literature. By associating the demand process with an underlying Markov chain, we are able to capture the effect of environmental factors that influence the demand process. Although the modeling capability is significantly enhanced by the incorporation of Markovian demands in inventory models, the simplicity of the optimal policies normally exhibited in the classical inventory problems is still preserved. Specifically, we show that the (s, S)-type policies shown to be optimal for a large class of inventory models with independent demands continue to be optimal for Markovian demand models, with one difference. That is, with Markovian demands, the (s, S) values depend on the state of the Markov process.
KeywordsOptimal Policy Inventory Model Markov Decision Process Demand Process Inventory Problem
Unable to display preview. Download preview PDF.