ABSTRACT
This chapter focuses on fitting cure rate models, which are mixture models in survival analysis, with integrated nested Laplace approximation (INLA). Multimodal data are not uncommon in statistics and they often appear when observations come from two or more underlying groups or populations. The analysis of these types of data is often undertaken with mixture models, which make use of different components that can model the different groups or populations in the data. The analysis of mixture models is complex. The chapter also provides a short link between INLA and these models. The computational approach to fit mixture models with INLA has the advantage of providing a toolbox for building different types of mixture models, with different types of likelihoods and priors on the hyperparameters not restricted to conjugate priors. Determining the number of components in a model is often difficult and a number of proposals have been made so far.
