Black Hole Discovery with Machine Learning

The Rubin Observatory's Legacy Survey of Space and Time (LSST) is a telescopic sky survey in Chile that is slated to start in the coming year. LSST will catalog 40 billion sources over the next ten years (with hundreds of repeat observations to track changes with time). The LSST's AGN Science Collaboration is charged with identifying which 100-300 million of those objects are so-called "Active Galactic Nuclei" (or "quasars"), which are galaxies whose central black holes are actively accreting new material. Identification of just 100 million objects out of 40 billion is very much a needles-in-a-haystack problem that is ideal for machine learning. I will discuss our approach to the problem and some of the challenges, including non-uniformity of the data (in both time and space). As AGNs seem to play a key active role in the evolution of galaxies and LSST will provide the largest data set to date, there is great potential to leverage these data for scientific discovery.