强混合误差下回归函数小波估计的渐近正态性

Asymptotic Normality of Wavelet Estimator of Regression Function under α-mixing Dependent Errors

文章讨论异方差非参数回归模型,在随机误差序列{ei,i≥1}为α-混合情形下,建立了回归函数g(.)的小波估计,并得到了该估计的渐近正态性,这些结果推广了梁汉营等人(2007年)在NA情形下的结论.

This paper discusses the heteroscedastic regression model. Under the random errors {ei, i ≥ 1} form a sequence of a-mixing random variables, the paper constructs the wavelet estimators of regression function g（ · ）, and establishes its asymptotic normality, which extends the results in NA framework investigated by Liang Hanying et al （2007）.