强相依数据的函数系数部分线性模型的估计  

Functional-Coefficient Partially Linear Models for Strong Mixing Dependent Data

作　　者：张日权[1] 

机构地区：[1]雁北师范学院数学系

出　　处：《应用数学学报》2006年第2期374-381,共8页Acta Mathematicae Applicatae Sinica

摘　　要：本文讨论在数据是强相依的情况下函数系数部分线性模型的估计．首先,采用局部线性方法,给出该模型函数项函数的估计;然后,使用两阶段方法给出系数函数的估计．并且讨论了函数项函数估计的渐近正态性,以及系数函数估计的弱相合性和渐近正态性.模拟研究显示,这些估计是较为理想的．In this paper, the functional-coefficient partially linear models are considered under data being strong mixing dependent. The local linear methods are used to estimate the functional part function. To estimate these coefficient functions, two-stage estimating process is employed. Under some conditions, the asymptotic normality of estimator of the functional part function is studied, as are the weak consistences and the asymptotic normalities of estimators of these coefficient functions. A simulated example is studied for illustration.

关 键 词：函数系数部分线性模型 强相依 局部线性方法 两阶段方法 弱相合性 渐近正态性 

分 类 号：O212.7[理学—概率论与数理统计]