Noisy Labels for Weakly Supervised Gamma Hadron Classification

TL;DR

This paper introduces a weakly supervised approach for gamma hadron classification in gamma ray astronomy that leverages noisy labels from real telescope data and uses detection significance as a learning criterion, eliminating the need for costly simulations.

Contribution

It presents a novel noisy label method for gamma hadron classification that achieves state-of-the-art results without relying on simulated ground-truth labels.

Findings

Models trained with noisy labels outperform traditional supervised methods.

The approach is effective on imbalanced datasets from various domains.

Only one class-wise noise rate needs to be known for the method to work.

Abstract

Gamma hadron classification, a central machine learning task in gamma ray astronomy, is conventionally tackled with supervised learning. However, the supervised approach requires annotated training data to be produced in sophisticated and costly simulations. We propose to instead solve gamma hadron classification with a noisy label approach that only uses unlabeled data recorded by the real telescope. To this end, we employ the significance of detection as a learning criterion which addresses this form of weak supervision. We show that models which are based on the significance of detection deliver state-of-the-art results, despite being exclusively trained with noisy labels; put differently, our models do not require the costly simulated ground-truth labels that astronomers otherwise employ for classifier training. Our weakly supervised models exhibit competitive performances also on imbalanced data sets that stem from a variety of other application domains. In contrast to existing work on class-conditional label noise, we assume that only one of the class-wise noise rates is known.