Neural topic models with survival supervision: Jointly predicting time-to-event outcomes and learning how clinical features relate

George H. Chen, Linhong Li, Ren Zuo, Amanda Coston, Jeremy C. Weiss

Abstract: Highlights•Deep survival models for predicting time-to-event outcomes are often uninterpretable•A new neural net framework is proposed for interpretable time-to-event prediction•The new framework learns how features relate in terms of a neural topic model•The resulting neural topic models predict as well as some deep survival baselines•Practical issues that arise in analyzing clinical data are discussed