Can semi-supervised learning reduce the amount of manual labelling required for effective radio galaxy morphology classification?

TL;DR

This paper investigates whether semi-supervised learning can reduce manual labeling in radio galaxy morphology classification, finding that SSL's effectiveness diminishes with very limited labels and unlabelled data.

Contribution

The study evaluates the robustness of SSL algorithms in radio astronomy and highlights their limitations with minimal labeled data and unlabelled datasets.

Findings

SSL offers regularisation benefits but degrades with fewer labels

Performance drops significantly with truly unlabelled data

SSL is less effective than supervised methods in low-label scenarios

Abstract

In this work, we examine the robustness of state-of-the-art semi-supervised learning (SSL) algorithms when applied to morphological classification in modern radio astronomy. We test whether SSL can achieve performance comparable to the current supervised state of the art when using many fewer labelled data points and if these results generalise to using truly unlabelled data. We find that although SSL provides additional regularisation, its performance degrades rapidly when using very few labels, and that using truly unlabelled data leads to a significant drop in performance.