In this paper, for the generalized estimating equation (GEE) with diverging number of covariates, the asymptotic properties of GEE estimator are considered. Under the weaker assumption on the minimum eigenvalue of Fisher information matrix and some other regular conditions, we prove the asymptotic existence, consistency and asymptotic normality of the GEE estimator and the asymptotic distribution of the test statistics of linear combination of the unknown parameters. The results are illustrated by Monte-Carlo simulations.

中文翻译：

---

具有协变量发散维数的 GEE 的渐近性质

在本文中，对于协变量数发散的广义估计方程（GEE），考虑了GEE估计量的渐近性质。在Fisher信息矩阵最小特征值等一些正则条件的较弱假设下，证明了GEE估计量的渐近存在性、一致性和渐近正态性，以及未知参数线性组合检验统计量的渐近分布。蒙特卡洛模拟说明了结果。