The eROSITA Final Equatorial-Depth Survey (eFEDS): A machine learning approach to inferring galaxy cluster masses from eROSITA X-ray images
Sven Krippendorf, Nicolás Barón Pérez, Esra Bülbül +9 more
We have developed a neural network-based pipeline to estimate masses of galaxy clusters with a known redshift directly from photon information in X-rays. Our neural networks were trained using supervised learning on simulations of eROSITA observations, focusing on the Final Equatorial Depth Survey (eFEDS). We used convolutional neural networks that have been modified to include additional information on the cluster, in particular, its redshift. In contrast to existing works, we utilized simulations that include background and point sources to develop a tool that is directly applicable to observational eROSITA data for an extended mass range – from group size halos to massive clusters with masses in between 10 13 M ⊙ < M < 10 15 M ⊙ . Using this method, we are able to provide, for the first time, neural network mass estimations for the observed eFEDS cluster sample from…