Events

In clinical practice, it is often useful to predict the probability that a subject will experience an event.  A Super Learner makes it possible to combine different regression models and algorithms. We have proposed the R package “survivalSL”, which contains various functions for building a Super Learner in the presence of censored event-time data. Compared with existing solutions, we provide a large number of learners and loss functions. We conducted simulations to describe the performance of our approach and also illustrated its use in an application to multiple sclerosis.

Keywords: Censored data, prediction, survival, machine learning, super learning

Stay tuned for all the information about the third edition of the PhD Student Award competition!