Reinterpreting survival analysis in the universal approximator age

TL;DR

This paper advances survival analysis by integrating deep learning, introducing a new loss function, evaluation metrics, and a universal approximator network that efficiently models survival curves.

Contribution

It provides a novel deep learning framework for survival analysis, including a new loss function, evaluation metrics, and a universal approximator network that avoids numeric integration.

Findings

The new loss function and model outperform existing approaches.

The universal approximator network effectively models survival curves.

The approach connects survival analysis with classification and regression.

Abstract

Survival analysis is an integral part of the statistical toolbox. However, while most domains of classical statistics have embraced deep learning, survival analysis only recently gained some minor attention from the deep learning community. This recent development is likely in part motivated by the COVID-19 pandemic. We aim to provide the tools needed to fully harness the potential of survival analysis in deep learning. On the one hand, we discuss how survival analysis connects to classification and regression. On the other hand, we provide technical tools. We provide a new loss function, evaluation metrics, and the first universal approximating network that provably produces survival curves without numeric integration. We show that the loss function and model outperform other approaches using a large numerical study.