ARC Project

 

A fundamental question that arises in any survival data analysis is the examination and modeling of the relationship between the time to event and one or more predictors. Estimating the survival function and related quantities like the conditional mean function and the conditional quantile function are very useful for the practitioners. However, due to censoring, one cannot use classical statistical techniques, that assume that the data are complete. The estimation and inference problem becomes even more challenging if the population contains an immune (cure) portion. This is because, immune individuals manifest themselves only by a large censored survival time. 
 

By combining our efforts and our knowledge and by providing innovative ideas and using recent developments in nonparametric and semiparametric analysis, we hope to make significant progress on the topic of cure models and more generally in the domain of survival analysis. Besides the methodological developments of this project, another objective will also be to apply these new results to real datasets.