缺失数据下半参数回归模型的局部线性光滑

Local linear smoothing to semiparametric regression model under missing response data

在缺失响应变量的不完全数据下,对半参数回归模型进行研究.利用最小二乘和局部线性回归拟合方法建立缺失数据下半参数回归模型参数分量和非参数分量的局部线性估计.在适当的条件下,得到^βn,^nσ的渐近正态性和^gn(t)最优弱收敛速度.

The semiparametric regression model was investigated for the case of response data missing. The local linearity estimations of parametric and nonparametric components in semiparametric regression model were made by using a method of local linear regressive fitting in combination with LS method. In an appropriate condition, the ^βn,^nσ of their asymptotic normalities and the ^gn（t） of optimal weak convergence rate were obtained.