Netherlands Astronomy Conference 2026

Images are among the most important tools astronomers use to understand the Universe. As machine learning techniques are increasingly used in astronomy, we explore what machines “see” in galaxy images, with a particular focus on galaxy morphology. 

Galaxy morphology provides one of the most direct observational signatures of evolutionary pathways, making its taxonomy a central pursuit in astronomy for over a century. Traditional classification schemes, most notably the Hubble sequence, rely heavily on visual inspection and human-defined categories. While foundational, these frameworks may not fully capture the diversity and complexity of structures revealed by modern imaging surveys. In this talk, I will present previous works applying machine learning techniques to the study of galaxy morphology from images, highlighting how data-driven approaches can uncover new perspectives on the structural diversity of galaxies beyond conventional classification schemes.