Using Machine Learning and Swift-XRT to Characterize likely X-ray Counterparts to Fermi Unassociated Sources

The Fermi Gamma Ray Observatory has revolutionized gamma-ray astronomy by discovering thousands of sources since its launch in 2008. However, the unidentified population of these sources in the Fermi catalogs is still substantial, e.g. one-third of the Fermi sources in the 3FGL catalog are unidentified. Swift-XRT observations of these Fermi unassociated fields have found possible X-ray counterparts in $\sim$30\% of these Fermi unassociated uncertainty regions, and approximately half of these sources were previously uncatalogued in either radio/optical/X-ray catalogs. The main objective of this work is to identify the nature of these possible counterparts, utilizing the properties of known Fermi sources coupled with the X-ray source properties. The majority of the known sources in the Fermi catalogs are blazars, which constitute the bulk of the extragalactic gamma-ray source population. The galactic population, on the other hand, is dominated by pulsars. Overall, these two categories constitute the majority of all gamma-ray objects. Blazars and pulsars occupy different parameter space when X-ray fluxes are compared with various gamma-ray properties. In our work, we utilize the X-ray observations performed with the Swift-XRT for the unknown Fermi sources and compare the X-ray and gamma-ray properties of possible X-ray associations in order to differentiate between blazars and pulsars. We then employ two machine learning algorithms to our high signal-to-noise ratio sample (121 X-ray sources that are likely to be associated with the gamma-ray source) identifying them with a pulsar or a blazar type. From this sample, we find that ~80% of the sources are likely blazars and ~3% are likely pulsars, at the 90% confidence interval. There are also a handful of new X-ray associations which are not clearly identified as either pulsars or blazars.