摘要
§1.引言和主要结果 设{(X_i,Y_i);1≤i≤n}为来自二元总体的iid样本。对于回归函数r(x)=E(Y|X=x)的估计问题,早在1964年。
In this paper,the kernel estimation of nonparametric regression functions is considered. We give a useful decomposition for the error of kernel estimation by means of Von Mises approach.In terms of this decomposition,we obtain a normal approximation rate of kernel estimation of regression functions on every finite interval.
出处
《应用数学学报》
CSCD
北大核心
1991年第3期304-311,共8页
Acta Mathematicae Applicatae Sinica
