# Survival Function Matching for Calibrated Time-to-Event Predictions

## TL;DR

This paper introduces a neural network-based survival function estimator for time-to-event predictions that improves calibration and distribution concentration without relying on adversarial training or explicit distribution assumptions.

## Contribution

It proposes a novel survival function estimator that calibrates probabilistic predictions without explicit distribution assumptions or adversarial training, enhancing calibration and distribution concentration.

## Key findings

- Outperforms existing models in calibration accuracy.
- Achieves better distribution concentration of predicted survival times.
- Works effectively without adversarial training.

## Abstract

Models for predicting the time of a future event are crucial for risk assessment, across a diverse range of applications. Existing time-to-event (survival) models have focused primarily on preserving pairwise ordering of estimated event times, or relative risk. Model calibration is relatively under explored, despite its critical importance in time-to-event applications. We present a survival function estimator for probabilistic predictions in time-to-event models, based on a neural network model for draws from the distribution of event times, without explicit assumptions on the form of the distribution. This is done like in adversarial learning, but we achieve learning without a discriminator or adversarial objective. The proposed estimator can be used in practice as a means of estimating and comparing conditional survival distributions, while accounting for the predictive uncertainty of probabilistic models. Extensive experiments show that the proposed model outperforms existing approaches, trained both with and without adversarial learning, in terms of both calibration and concentration of time-to-event distributions.