Dr. Reinaldo de Carvalho, (Universidade de Sao Paulo)
21 Maggio 2024
11:30, Aula Piazzi
Abstract:
Galaxy morphology offers significant insights into the evolutionary pathways and underlying physics of galaxies. As astronomical data grow with surveys such as CEERS/JWST, there is a need for tools to classify and analyse the vast numbers of galaxies that will be observed. In this presentation I will introduce a novel classification technique blending unsupervised clustering based on morphological metrics with the scalability of supervised Convolutional Neural Networks. We present a comparative analysis between the well-known CAS (Concentration, Asymmetry, and Smoothness) metrics and our newly proposed EGG (Entropy, Gini, and Gradient Pattern Analysis). Our choice of the EGG system stems from its separation-oriented metrics, maximizing morphological class contrast. Applying our methodology to the Sloan Digital Sky Survey data, we obtain ∼95% of Overall Accuracy of purely unsupervised classification and when we replicate T-Type and visually classified galaxy catalogues with accuracy of ∼88% and ∼89%, respectively. Furthermore, the application to Hubble Space Telescope data heralds the potential for unsupervised exploration of a higher redshift range. A notable achievement is our ∼95 per cent accuracy in unsupervised classification, a result that rivals when juxtaposed with Traditional Machine Learning and closely trails when compared to Deep Learning benchmarks. Some preliminary results using CANDELS data will be presented and discussed.