Classification of Fermi-LAT unassociated sources with machine learning in the presence of dataset shifts
arXiv:2412.04675 · doi:10.1051/epjconf/202531906009
Abstract
About one third of Fermi Large Area Telescope (LAT) sources are unassociated. We perform multi-class classification of Fermi-LAT sources using machine learning with the goal of probabilistic classification of the unassociated sources. A particular attention is paid to the fact that the distributions of associated and unassociated sources are different as functions of source parameters. In this work, we address this problem in the framework of dataset shifts in machine learning.
3 figures, 4 pages, 5 equations; proceeding from Roma International Conference on AstroParticle Physics (RICAP-24)