基于分位数自适应LASSO方法线性面板模型的变点估计  

Change Point Estimation in Linear Panel Model ViaQuantile Adaptive LASSO Method

作　　者：陈小红 齐培艳 CHEN Xiao-hong;QI Pei-yan(School of Applied Sciences,Taiyuan University of Science and Technology,Taiyuan 030024,China)

机构地区：[1]太原科技大学应用科学学院,太原030024

出　　处：《太原科技大学学报》2025年第1期64-70,共7页Journal of Taiyuan University of Science and Technology

基　　金：山西省基础研究计划(202103021224274);山西省社会经济统计科研课题(KY[2021]145)。

摘　　要：采用分位数自适应LASSO方法针对线性面板数据模型中的变点估计问题进行了研究,给出了变点位置、个数及模型参数的估计,并推导了该方法下变点个数估计及参数估计的一致性。数值模拟验证了两种误差分布下分位数自适应LASSO方法的估计性能优于分位数LASSO方法;最后通过应用居民消费价格总指数数据说明了分位数自适应LASSO方法的有效性。The quantile adaptive LASSO method is used to study the changepoints estimation in linear panel data model,the estimation of change point position,number and model parameters are given,and the consistency of change point number estimation and parameter estimation under this method is deduced.Numerical simulation shows that the estimation performance of quantile adaptive LASSO method is better than that of quantile LASSO method under two error distributions;Finally,the effectiveness of quantile adaptive LASSO method is illustrated by applying the data of general consumer price index.

关 键 词：分位数自适应LASSO方法 线性面板模型 变点估计 

分 类 号：O212.1[理学—概率论与数理统计]