Mixture cure models: a goodness-of-fit perspective

Abstract

Here we revisit a goodness-of-fit testing problem for randomly right-censored data in the presence of a cured subject. The focus of the talk is on the modification of recently proposed characterization-based goodness-of-fit tests for exponential distribution constructed via IPCW U- or V-approach. We present their asymptotic properties and extend our discussion to encompass suitable generalizations applicable to a variety of tests formulated using the same methodology. We conduct a comparative analysis of these proposed tests against a recent competitor based on the CvM distance. Additionally, we explore modifications of the most prominent competitors identified in prior studies that didn't consider the presence of cured subjects.