Bias Correction for a Generalized Log-Gamma Regression Model

Deviance residuals in generalised log-gamma regression models with censored observations

“…Email: giapaula@ime.usp.br survival analysis and reliability and has as specific cases other models broadly used in the lifetime data analysis. For instance, DiCiccio [2] derives approximate inferences for the quantiles and scale parameters whereas Young and Bakir [3] obtain the bias of order n −1 , where n is the sample size, for the parameter estimates in generalised log-gamma regression models for uncensored samples.Young and Bakir [3] also present the expectation of various log-likelihood derivatives in closed-form expressions. Ahn [4] proposes a regression tree method to clarify the heterogeneous subsets of the data into different generalised log-gamma regression models with the shape parameter being estimated separately in each formed stratum under independent random censoring and Ortega et al [5] derive the appropriate matrices for assessing local influence on the parameter estimates under different perturbation schemes.…”

Bias correction for exponential family nonlinear modles

“…Also for this class of models, Cook et al (1986) show that the bias may be due to the explanatory variables position in the sample space. Young & Bakir (1987) use bias correction to improve several pivotal random variables for a generalized log-gamma regression model. Recently, Cordeiro & McCullagh (1991) gave a general bias formulae in matrix notation for generalized linear models (McCullagh & Nelder, 1989).…”

Improved maximum-likelihood estimation in a regression model with general parametrization

“…For example, Cook et al [7] proposed bias correction in normal nonlinear models. In [8], bias-corrected estimates for a generalized log-gamma regression model are obtained. In [9] a general matrix formulae for bias correction in generalized linear models is given, whereas Paula [10] derived the second-order biases in exponential family nonlinear models.…”