A Review of Semi-parametric Models for Life Histories

Abstract

The development of various generalizations of Cox's proportional hazards regression model to the study of more complicated life histories is reviewed. Cox's paper provided a framework for the regression modelling of single failure (deaths) in terms of a proportional hazards assumption and subsequent work has extended this approach to the cases of multiple (non-catastrophic) failures, multi-state life histories, non-independent failure times (family studies and matched pairs), and more complicated censoring mechanisms.