Logistic回归中的加权最小二乘估计

The Weighted Least Estimation of Logistic Models

本文引入因变量是虚拟变量的回归模型 ,并通过logistic函数拟合这种模型 ,分析它在控制和预测方面的不足之处 .在此基础上提出用加权最小二乘估计回归方程 .模拟结果表明 。

This paper introduces the logistic regression models.First it extends a kind of models that the dependent variable is dichotomous.Then estimates it through logistic function.But there are some shortcomings when predicting.We give the weighted least estimation based on the heteroscedasticity.