The Power Series Cure Rate Model: An Application to a Cutaneous Melanoma Data

Abstract

In this article we propose a new cure rate survival model. In our approach the number of competing causes of the event of interest is assumed to follow an exponential discrete power series distribution. An advantage of our model is that it is very flexible, including several particular cases, such as, Bernoulli, geometric, Poisson, etc. Moreover, we derive the appropriate matrices for assessing local influence on the parameter estimates under different perturbation schemes and present some ways to perform global influence analysis. Distribution fitting can be tested for the best fitting in a straightforward way. Maximum likelihood estimation is discussed. Our proposed model is illustrated through cutaneous melanoma data.

Keywords:

• Competing Risks
• Cure rate models
• Power series distribution

Mathematics Subject Classification:

• 62Nxx

Acknowledgment

The research was granted by CNPq-Brasil.