Highly robust and efficient estimators for generalized linear models with a dispersion parameter are proposed. The estimators are based on three steps. In the first step, the maximum rank correlation estimator is used to consistently estimate the slopes up to a scale factor. The scale factor, the intercept, and the dispersion parameter are robustly estimated using a simple regression model. Then, randomized quantile residuals based on the initial estimators are used to define a region S such that observations out of S are considered as outliers. Finally, a conditional maximum likelihood (CML) estimator given the observations in S is computed. We show that, under the model, S tends to the whole space for increasing sample size. Therefore, the CML estimator tends to the unconditional maximum likelihood estimator and this implies that this estimator is asymptotically fully efficient. Moreover, the CML estimator maintains the high degree of robustness of the initial one. The negative binomial regression case is studied in detail.
Generalized linear model Conditional maximum likelihood Negative binomial regression Overdispersion Robust regression
Mathematics Subject Classification
62F10 62F12 62F35 62J12 62J20
This is a preview of subscription content, log in to check access.
Cantoni E, Zedini A (2009). A robust version of the hurdle model. Cahiers du département d’économétrie No 2009.07, Faculté des sciences économiques et sociales, Université de GenèveGoogle Scholar
Carter EM, Potts HWW (2014) Predicting length of stay from an electronic patient record system: a primary total knee replacement example. BMC Med Inform Decis Mak 14:26CrossRefGoogle Scholar
Cuesta-Albertos JA, Matrán C, Mayo-Iscar A (2008) Trimming and likelihood: robust location and dispersion estimate in the elliptical model. Ann Stat 36(5):2284–2318CrossRefGoogle Scholar
Davison AC, Snell EJ (1991) Residuals and diagnostics. In: Hinkley DV, Reid N, Snell EJ (eds) Statistical theory and modelling: in honour of Sir David Cox. Chapman and Hall, Boca Raton, pp 83–106Google Scholar
Dunn PK, Smyth GK (1996) Randomized quantile residuals. J Comput Graph Stat 5(3):236–244Google Scholar
Ghosh A, Basu A (2013) Robust estimation for independent non-homogeneous observations using density power divergence with applications to linear regression. Electron J Stat 7:2420–2456MathSciNetCrossRefGoogle Scholar
Künsch HR, Stefanski LA, Carroll RJ (1989) Conditionally unbiased bounded-influence estimation in general regression models, with applications to generalized linear models. J Am Stat Assoc 84(406):460–466MathSciNetzbMATHGoogle Scholar