由區域線性迴歸平滑方法來估計改變點位置及改變量

Abstract

In the fixed design nonparametric regression model, we propose a change-point estimator based on maximizing the difference of two one-sided local linear regression smoothers and the corresponding jump size estimator being the difference at the change-point estimator. The asymptotic normality is established for both estimators under regular conditions. We also propose two estimators for the underlying regression function. A simulation is conducted to demonstrate the performance of the method in finite sample situations. The method is also applied to one famous data set, Nile data.