DSIG Seminar: Neural Networks for Time-to-Event Data Analysis

Recent recommendations advocate for classical machine learning and deep learning approaches in survival probability prediction. Additionally, a novel neural network architecture, rooted in nested case-control methodology, has emerged. This architecture demonstrates scalability to large datasets and accommodates both proportional and non-proportional extensions of the Cox model for single-event analysis. In this research study, we build a neural network based on this neural network architecture for a single event analysis.