Abstract. In this paper, a simplified parsimonious higher-order multivariate Markov chain model (SPHOMMCM) is presented. Moreover, parameter estimation method of TPHOMMCM is give. Numerical experiments shows the effectiveness of TPHOMMCM. Several prediction models of multiple categorical data sequences has been exhibited, e.g., the firstorder multivariate Markov chain model, the higher-order multivariate Markov chain model and the improved multivariate Markov chain model (which adds a negative association part at the back of the normal model.)
IntroductionIn this article, a simplified parsimonious higher-order multivariate Markov chain model is proposed for a better the prediction results and less computational cost.The organization of this paper is as follows. In Section 2, we review several basic knowledge of Markov chain model. In Section 3, a simplified parsimonious higher-order multivariate Markov chain model is presented for different data sequences. In Section 4, the parameters estimation method of SPHOMMCM is given. Finally, numerical experiments illustrate the effectiveness of our model in Section 5.