强混合序列下非参回归函数加权核估计的强收敛速度

Strong Convergence Rate of Weighted Kernel Estimator of Nonparametric Regression Functions under Strong Mixing Sequences

在误差项为强混合序列的条件下,利用随机变量部分和的矩不等式,讨论非参回归函数加权核估计的强相合性,给出其收敛速度.当样本矩足够大时,强相合的收敛速度约等于n-1/2.

The strong consistency of weighted kernel estimator of nonparametric regression functions is discussed under strong mixing errors, and its convergence rate is also given. As the moments of samples is large enough,the convergence rate equals approximately to n^-1/2 according to the theorem.