STATISTICAL ANALYSIS OF SURVIVAL TIMES BASED ON PROPORTIONAL GENERALIZED ODDS MODELS
Keywords:
Burr distribution, Frailty parameter, Maximum likelihood estimation, Censored dataAbstract
The Cox proportional hazards (PH) model is the most widely used regression model in the field of survival analysis. Unfortunately, the PH model cannot be used in reality since not all data sets meet the PH criteria. The proportional odds (PO) model (Pettitt 1982 and Bennett 1983a) and the generalized proportional odds (GPO) model (Dabrowska and Doksum, 1988), which are essentially extensions of the PH model, were offered as solutions to the issue. Nevertheless, there are instances that suggest that the PO or GPO paradigm is inappropriate. Consequently, a broader model has to be taken into account. The proportional generalized odds (PGO) model, which includes the PO and GPO models as specific examples, is presented in this study. The GPO model's underlying survival function and regression parameter estimation are examined. The model is applied to a collection of data.