Robust Inference in Generalized Linear Models

被引:8
|
作者
Alqallaf, Fatemah [1 ]
Agostinelli, Claudio [2 ]
机构
[1] Kuwait Univ, Dept Stat & Operat Res, Jamal Abdul Nasser St, Safat, Kuwait
[2] Univ Trento, Dept Math, Trento, TN, Italy
关键词
Gamma model; Inverse Gaussian model; Outliers; Poisson model; Robust estimation; Robust model selection; Weighted likelihood; Primary; 62F35; 62FG35; Secondary; 62-07; WEIGHTED LIKELIHOOD METHODOLOGY; LOGISTIC-REGRESSION MODEL; ESTIMATING EQUATIONS; RESIDUALS; DEFINITION; ESTIMATOR; TESTS;
D O I
10.1080/03610918.2014.911896
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
Robust inference on the parameters in generalized linear models is performed using the weighted likelihood method. Two cases are considered: a case with replicated observations and a case with a single observation of the dependent variable for each combination of the explanatory variables. The first case is common in the design of experiments, while the second case arises in observational studies. Theoretical and computational results on real datasets are presented and compared with other existing techniques.
引用
收藏
页码:3053 / 3073
页数:21
相关论文
共 50 条