Analysis of breast cancer data in framework of a GPD model with interval censoring

Abstract

In this work, we are interested in a hypothesis testing problem within the framework of a GPD model with interval censoring. For this purpose, we rst develop the calculation of the likelihood function using conditional probabilities to achieve the same expression proposed by Klein and Moeschberger. Next, we show that the properties of the maximum pseudo-likelihood estimates of the model parameters, and essentially the asymptotic normality, are preserved. Finally, we built a hypothesis testing to compare two types of breast cancer treatment as part of the model mentioned above.