DeepMerge II: Building Robust Deep Learning Algorithms for Merging Galaxy Identification Across Domains

TL;DR

This paper demonstrates that domain adaptation techniques significantly improve the accuracy of deep learning models in classifying merging galaxies across simulated and real astronomical datasets, enabling better cross-domain generalization.

Contribution

The paper introduces the application of domain adaptation methods like MMD and DANNs to astronomy, improving neural network robustness across different data domains for galaxy classification.

Findings

Domain adaptation techniques increase classification accuracy by up to 20%.

Methods work effectively on both simulated and real astronomical data.

Enhanced domain-invariant feature extraction improves model transferability.

Abstract

In astronomy, neural networks are often trained on simulation data with the prospect of being used on telescope observations. Unfortunately, training a model on simulation data and then applying it to instrument data leads to a substantial and potentially even detrimental decrease in model accuracy on the new target dataset. Simulated and instrument data represent different data domains, and for an algorithm to work in both, domain-invariant learning is necessary. Here we employ domain adaptation techniques$-$ Maximum Mean Discrepancy (MMD) as an additional transfer loss and Domain Adversarial Neural Networks (DANNs)$-$ and demonstrate their viability to extract domain-invariant features within the astronomical context of classifying merging and non-merging galaxies. Additionally, we explore the use of Fisher loss and entropy minimization to enforce better in-domain class discriminability. We show that the addition of each domain adaptation technique improves the performance of a classifier when compared to conventional deep learning algorithms. We demonstrate this on two examples: between two Illustris-1 simulated datasets of distant merging galaxies, and between Illustris-1 simulated data of nearby merging galaxies and observed data from the Sloan Digital Sky Survey. The use of domain adaptation techniques in our experiments leads to an increase of target domain classification accuracy of up to ${\sim}20\%$. With further development, these techniques will allow astronomers to successfully implement neural network models trained on simulation data to efficiently detect and study astrophysical objects in current and future large-scale astronomical surveys.