一类零膨胀广义线性模型极大似然估计的相合性与渐近正态性

Consistency and Asymptotic Normality of Maximum Likelihood Estimation for Zero-inflation Generalized Linear Models

讨论了响应变量为单参数指数族且在零点处膨胀的广义线性模型的大样本性质,对其参数进行了极大似然估计,给出了一些正则条件.基于所提出的正则条件,证明了模型参数极大似然估计的相合性与渐近正态性.

The large sample nature of the generalized linear model with a single-parameter exponential family and inflation at zero is discussed.The maximum likelihood estimation is performed on its parameters.Some regular conditions are given.Based on the proposed regular conditions,the model is proved consistency and asymptotic normality of parameter maximum likelihood estimation.