Some issues on inference for a class of univariate frailty regression models

Abstract

The frailty model generalizes the usual proportional hazards model by incorporating a random effect. Estimation in such model has received much attention in recent years. EM and penalized likelihood method have already employed to estimate the parameters in certain frailty models like Gamma and Log-Normal. The corresponding consistency and asymptotical theory have been already been proved. This thesis addresses how to estimate parameters using MCMC method in Inverse Gaussian model and provide the consistency proof. Also there is a comprehensive simulation aims to study the robustness of the proposed method under misspecification.