多元非线性回归模型GLS估计的渐近性质

Asymptotical properties of generalized least squares estimation for multiple nonlinear regression models

文章给出了多元非线性回归模型的广义最小二乘法(GLS)估计,并运用Klimko-Nelson定理关于随机过程的OLS估计量是强相合和渐近正态的结果,证明了多元非线性回归模型的GLS估计量的强相合性、渐近正态性和渐近有效性。

This paper presents the estimator of GLS of parameters for the multiple nonlinear regression models. By using the Klimko-Nelson theorem, in which the OLS estimator of a random process has the results of strong consistency and asymptotical normal, it is proved that the GLS estimator has characteristics of strong consistency, asymptotical normal and asymptotical efficiency.