Multivariate Markov chain models for production planning

Authors:
  • DZDong Mei Zhu
    Dong Mei Zhu
  • WCWai Ki Ching
    Wai Ki Ching

Abstract
Markov models are commonly used in modelling many practical systems such as queueing networks, manufacturing systems and inventory systems. In this paper, we consider a multivariate Markov chain model for modelling multiple categorical data sequences. We develop new efficient estimation methods for the model parameters. We then apply the model and methods to demand predictions and production planning for a soft-drink company in Hong Kong. This problem is essentially a newsboy's problem in a multivariate Markov chain framework. Numerical examples are given to demonstrate the effectiveness of the proposed methods.