Statistical Analysis Of Parameter Multiple Change Points In Normal Distribution

The problem of change point has always been one of the focus topics in statistics,which was put forward from the quality control process of industrial production.With the in-depth study of the problem,the statistical inference of change point has been applied to many fields such as economy,finance,biomedicine,meteorology,geology and so on.In multivariate statistical inference,it is not only necessary to estimate the number of change points,but also to estimate the position of the change points.Generally,the two are interdependent.The main task of this paper is to estimate the number and position of the multivariate points of the parameters of the normal population.Inspired by the binary segmentation method and the segmentation method in the wild binary segmentation method,a new sequential segmentation method is proposed for the sequence segmentation.Inspired by the binary segmentation method and the segmentation method in the wild binary segmentation method,a new sequential segmentation method is proposed for sequence segmentation,and the sequential segmentation method is introduced into the estimation of the number and position of the mean and variance points of the normal population,which is called the sequential segmentation estimation method.In this paper,the realization process of sequential piecewise estimation method is given,and the consistency of variable point estimation is proved.Finally,the validity of the sequential segmentation estimation method is verified by numerical simulations,and the sequential segmentation estimation method of the mean change points is applied to estimate the number and location of the mean change points in the railway passenger volume data.The simulation results show that the proposed method has good applicability.