ρ混合序列下非参回归函数加权核估计的强收敛速度

Strong Consistency Rate of Weighted Kernel Estimators of Nonparametric Regression Functions Under ρ Mixing Sequence

在误差项为ρ混合序列的条件下,讨论了非参回归函数加权核估计的强相合性,并给出其收敛速度.所得定理表明,当样本矩r充分大时,强相合的收敛速度约等于n-1/2.

Under the condition of the error term being a ρ mixing sequence, the strong consistency of weighted kernel estimators of nonparametric regression functions is discussed, and the convergence rate is given. The theorem provided in the paper shows that the convergence rate of strong consistency approxi-mately equals to when the sample moment r is large enough.