ρ-混合误差下回归函数估计的渐近正态

A symptotical Normality of the KernelEstimator for Regression Function Under d -Mixing Errors

设Y1,Y2,…,Yn 是固定点x1,x2,…,xn 的n 个观察值,适合模型Yi= g(xi)+ εi,1≤i≤n.在｛εi｝为ρ混合误差下讨论了Priestley 等人提出的一类非参数回归函数加权核估计的渐近正态性;在较弱条件下,通过对统计量分块的方法。

Let Y1 ,…,Yn be n observations at fixed points x1,…,xn according to the m odel yi= g (xi)+ εi,1≤i≤n,discuss the estim ator for a nonparam etric regression function given by priestley, get the asym ptotical norm ality of the estim ator under w eaker conditions.