固定情形Logistic回归模型误差方差的经验似然估计  

Empirical Likelihood Inference of Error Variance in Logistic Regression Models with Fixed Design Case

作　　者：胡学叶 张正家 HU Xueye;ZHANG Zhengjia(College of Mathematics and Statics,Guangxi Normal University,Guilin 541004,China)

机构地区：[1]广西师范大学数学与统计学院,广西桂林541004

出　　处：《广西民族大学学报（自然科学版）》2020年第2期58-63,共6页Journal of Guangxi Minzu University ：Natural Science Edition

摘　　要：应用经验似然方法得到了固定情形Logistic回归模型误差方差的经验似然估计,结果表明其渐近方差比传统残差平方和方法得到的估计更小.In this paper,we use the empirical likelihood method to obtain an empirical likelihood esti-mator for the error variance of the logistic regression models with fixed design.As a result,the asymptotic variance of the estimator given in this paper is less than the traditional one,which is the sum of squares of residuals.

关 键 词：经验似然 LOGISTIC回归模型 误差方差 残差平方和 

分 类 号：O212.7[理学—概率论与数理统计]