夏志明Zhiming Xia

教授

 博士生导师  硕士生导师
性别:男
学历:博士研究生毕业
在职信息:在职
所在单位:数学学院
电子邮箱:

论文成果

当前位置: 中文主页 >> 论文成果

Multi-stage Estimation for Mean-shift Model with the Unknown Number of Change-points

点击次数:

影响因子:0.0
发表刊物:Acta Mathematicae Applicatae Sinica
关键字:mean-shift model; local-comparison method; change-points; threshold value
摘要:Abstract In this paper, based on local comparison, a multi -stage estimation method for the number and the location of change-points is proposed in mean- shift model with unknown multiple change points. First, the alarm interval sequence of possible change point positions are estimated; then the number and the location of change points are prelirminarily estimated according to the alarm intervals; finally, the false change points are deleted to determine the final number and the location of change points. The theoretical results show that almost surely when the n is sufficiently large, the number of change points will be exactly equal to the true number of the change points, and the final alarm interval will strictly contain all true change-points. The estimator has strong consistency. The simulation results show that the multi-stage estimation can accurately estimate the number and the location of change- points under different conditions.
论文类型:期刊论文
卷号:40
期号:6
页面范围:931-952
ISSN号:0254-3079
是否译文:
第一作者:Zhiming Xia
通讯作者:Zhiming Xia