Title :
A New Estimation Method for Multivariate Markov Chain Model with Application in Demand Predictions
Author :
Zhu, Dong-Mei ; Ching, Wai-Ki
Author_Institution :
Dept. of Math., Univ. of Hong Kong, Hong Kong, China
Abstract :
In this paper, we propose a new estimation method for the parameters of a multivariate Markov chain model. In the new method, we calculate the correlations of the sequences first and establish multivariate Markov chain models for those positively correlated sequences. The parameters are estimated by minimizing the error of prediction. We apply the method to demand predictions for a soft-drink company in Hong Kong. Numerical experiments are given to show the effectiveness of our proposed method.
Keywords :
Markov processes; demand forecasting; minimisation; parameter estimation; Hong Kong; correlated sequences; demand predictions; error prediction minimization; multivariate Markov chain model; parameter estimation method; soft-drink company; Biological system modeling; Companies; Data models; Marketing and sales; Markov processes; Numerical models; Predictive models; Demand Prediction; Multivariate Markov Chain Model;
Conference_Titel :
Business Intelligence and Financial Engineering (BIFE), 2010 Third International Conference on
Conference_Location :
Hong Kong
Print_ISBN :
978-1-4244-7575-9
DOI :
10.1109/BIFE.2010.39
