ABSTRACT
Semiparametric and nonparametric cure models are proposed to relax the assumptions so that the results are less sensitive to the model assumptions. This chapter introduces semiparametric mixture cure models, which do not make parametric assumptions in the baseline distribution of the latency submodel. It introduces fully nonparametric approaches and presents semiparametric estimation methods for some non-mixture cure models. The chapter discusses some model assessment methods for semiparametric and nonparametric cure methods on available software for the models and their applications to real data. It primarily focuses on independent and right-censored data. The chapter shows some software packages and uses them to illustrate the applications of some cure models to real data sets.
