A Proper Scoring Rule for Validation of Competing Risks Models

TL;DR

This paper introduces a new proper scoring rule tailored for validating competing risks models, ensuring accurate evaluation of probabilistic predictions even with censored data.

Contribution

The authors propose a modified logarithmic scoring rule specifically designed for competing risks data, maintaining strict propriety under non-informative censoring.

Findings

The scoring rule remains strictly proper with censored data.

It provides a reliable method for model validation in competing risks scenarios.

The approach enhances existing evaluation techniques for survival analysis models.

Abstract

Scoring rules are used to evaluate the quality of predictions that take the form of probability distributions. A scoring rule is strictly proper if its expected value is uniquely minimized by the true probability distribution. One of the most well-known and widely used strictly proper scoring rules is the logarithmic scoring rule. We propose a version of the logarithmic scoring rule for competing risks data and show that it remains strictly proper under non-informative censoring.