Transfer learning for galaxy morphology from one survey to another

TL;DR

This study demonstrates that deep learning models trained on one galaxy survey can be effectively adapted to another survey with minimal additional data, significantly reducing the need for large labeled datasets in new surveys.

Contribution

It shows that transfer learning and domain adaptation enable accurate galaxy morphology classification across different surveys with fewer labeled examples.

Findings

Direct application yields ~90% accuracy with low completeness and purity.

Domain adaptation improves accuracy to >95% and enhances completeness and purity.

Significant reduction in required training samples for new survey data.

Abstract

Deep Learning (DL) algorithms for morphological classification of galaxies have proven very successful, mimicking (or even improving) visual classifications. However, these algorithms rely on large training samples of labelled galaxies (typically thousands of them). A key question for using DL classifications in future Big Data surveys is how much of the knowledge acquired from an existing survey can be exported to a new dataset, i.e. if the features learned by the machines are meaningful for different data. We test the performance of DL models, trained with Sloan Digital Sky Survey (SDSS) data, on Dark Energy survey (DES) using images for a sample of $\sim$5000 galaxies with a similar redshift distribution to SDSS. Applying the models directly to DES data provides a reasonable global accuracy ($\sim$ 90%), but small completeness and purity values. A fast domain adaptation step, consisting in a further training with a small DES sample of galaxies ($\sim$500-300), is enough for obtaining an accuracy > 95% and a significant improvement in the completeness and purity values. This demonstrates that, once trained with a particular dataset, machines can quickly adapt to new instrument characteristics (e.g., PSF, seeing, depth), reducing by almost one order of magnitude the necessary training sample for morphological classification. Redshift evolution effects or significant depth differences are not taken into account in this study.