Classification of X-ray sources from AXIS GPS and the population of extended high-energy sources

I will provide a quick overview of the supervised machine-learning approach to X-ray source identification and classification which relies on multiwavelength information. Automated classification methods are needed for sensitive X-ray surveys covering large areas on the sky, such as AXIS GPS, because of the large number of sources to be detected. I will also discuss challenges specific to the crowded region of the Galactic plane selected for the AXIS GPS. In addition to point sources, AXIS GPS is expected to discover new extended high-energy sources. I will provide a brief overview and specific examples of such sources and argue that deeper AXIS observations of some of the known extended sources would be a great supplement to the compact object science which is one of the main drivers behind GPS.