摘要
设{yi}是固定在点{xi}的观察值,适合模型yi=g(xi)+εi.其中g(x)是0,1上的未知函数,{εi}是均值为0的随机误差序列.文献中,在{εi}为独立同分布的条件下,通过构造新的函数gn(x),对g(x)进行了估计.论文将{εi}推广至~ρ-混合误差序列的情形,通过附加适当的条件和精细的计算,获得了用gn(x)估计g(x)的同样结论.
Let the observed process {yi} follow the regression model of the form yi= g（xi）＋εi, where {εi} is fixed on [0,1] and the sequence {εi} of errors is a stationary process with mean 0,and g（x） is an unknown function. In references,in the case of i. i. d. errors, some authors obtained an estimation for g （x） by constructing a new function gn（x）. This paper considers the ρ^^-mixing error and obtains the same result by some careful calculation under suitable conditions.
出处
《高校应用数学学报（A辑）》
CSCD
北大核心
2006年第4期439-444,共6页
Applied Mathematics A Journal of Chinese Universities(Ser.A)
关键词
ρ^^-混合
加权核估计
强相合
ρ^^-mixing
weighted kernel estimate
strong consistency
