Application of Hidden Markov and bayes for demand forecasting

Abstract

In applied forecasting one typically has under consideration not a single model but several alternative models. For choosing suitable model or suitable combination of models, some papers used the Bayesian approach or Hidden Markov Expert. This paper presents one approach for forecasting by combining Hidden Markov Model and Bayesian method in dealing with multi models in order to enhance the advantages of both methods © 2006 IEEE.