统计与决策2025，Vol.41Issue(19)：35-41,7.DOI:10.13546/j.cnki.tjyjc.2025.19.006
基于移动窗框方法的相关系数变点检验
Test of Change Point in Correlation Coefficient Based on the Moving Window Method
摘要
Abstract
Regarding the problem of change points in time series correlation coefficients,this paper uses the moving window method to convert the change points of correlation coefficients into mean change points.In order to avoid the consistent estimation of long-term variances,a modified Ratio statistic is proposed to test the mean change point,and the limit distribution of the statis-tic under the null hypothesis and the consistency under the alternative hypothesis are derived.At the same time,the consistent es-timator of the change point position is given.Due to the complexity of the limit distribution,Bootstrap sampling is applied to ap-proximate it and determine precise critical values.Simulation results indicate that the Bootstrap-based Ratio test maintains a good empirical size and demonstrates adequate empirical power.Finally,the validity and practicality of the proposed method are further confirmed by using two sets of stock market data.关键词
移动窗框/相关系数变点/均值变点/Ratio统计量/Bootstrap方法Key words
moving window/change point in correlation coefficient/mean change point/Ratio statistic/Bootstrap method分类
数理科学引用本文复制引用
孙治国,金浩,苏梦琳..基于移动窗框方法的相关系数变点检验[J].统计与决策,2025,41(19):35-41,7.基金项目
国家自然科学基金资助项目(71473194) （71473194）
陕西省科技厅自然科学基金资助项目(2020JM513) （2020JM513）
