Deployment of finite-stage Markov decision processes for inventory management solutions in enterprise restructuring, Chapter 13

摘要

The decision making process is becoming very complex, especially when it comes toudthe problem of enterprise restructuring. Intelligent solutions have to be utilized inudorder to cope efficiently with this issue. For this purpose the model for enterpriseudrestructuring, named as COMPASS was developed. It aims to systematise the complexudprocess of enterprise restructuring and to attach appropriate methods in the keyuddecision making points, supporting the industry praxis with robust, scientificallyudfounded, but practical and easy to use methods/tools. COMPASS is meant to be underudcontinuous development. Now, the improvement of the process of generation ofudsuccess factors (as a dominantly heuristic activity) and the optimization of theudsimulation process are the focal points. Here, the improvement of the generation ofudsuccess factors through an application of finite-stage Markov decision processesud(MDP) in the inventory problem is introduced. The objective is to find the optimaludinventory policy and to determine the utility function. MDP are a flexible technique forudstochastic and dynamic optimization problems and provide a powerful tool forudoptimizing performance in very different areas and their application in inventoryudmanagement is only one of the examples.