The generalized exponential-power series cure rate model with covariates and latent activation schemes.
Título da Revista
ISSN da Revista
Título de Volume
This report presents a new family of cure rate models derived from a scenario with latent risk factors under different activation schemes that would lead to the occurrence of the event of interest. It is considered that the times of activation of the latent factors follow a generalized exponential distribution, while the number of this factors follows a power series series distribution. These models incorporate explanatory variables related with the cure rate through logistic regression. A sensitivity analysis including diagnostic measures based on case-deletion approaches and local influence is performed. To illustrate the results of this paper an example with real data is presented.