Probabilistic classification of X-ray sources applied to Swift-XRT and XMM-Newton catalogs

Abstract

Context. Serendipitous X-ray surveys have proven to be an efficient way to find rare objects, for example tidal disruption events, changing-look active galactic nuclei (AGN), binary quasars, ultraluminous X-ray sources, and intermediate mass black holes. With the advent of very large X-ray surveys, an automated classification of X-ray sources becomes increasingly valuable. Aims. This work proposes a revisited naive Bayes classification of the X-ray sources in the Swift-XRT and XMM-Newton catalogs into four classes – AGN, stars, X-ray binaries (XRBs), and cataclysmic variables (CVs) – based on their spatial, spectral, and timing properties and their multiwavelength counterparts. An outlier measure is used to identify objects of other natures. The classifier is optimized to maximize the classification performance of a chosen class (here XRBs), and it is adapted to data mining purposes. Methods. We augmented the X-ray catalogs with multiwavelength data, source class, and variability properties. We then built a reference sample of about 25 000 X-ray sources of known nature. From this sample, the distribution of each property was carefully estimated and taken as reference to assign probabilities of belonging to each class. The classification was then performed on the whole catalog, combining the information from each property. Results. Using the algorithm on the Swift reference sample, we retrieved 99%, 98%, 92%, and 34% of AGN, stars, XRBs, and CVs, respectively, and the false positive rates are 3%, 1%, 9%, and 15%. Similar results are obtained on XMM sources. When applied to a carefully selected test sample, representing 55% of the X-ray catalog, the classification gives consistent results in terms of distributions of source properties. A substantial fraction of sources not belonging to any class is efficiently retrieved using the outlier measure, as well as AGN and stars with properties deviating from the bulk of their class. Our algorithm is then compared to a random forest method; the two showed similar performances, but the algorithm presented in this paper improved insight into the grounds of each classification. Conclusions. This robust classification method can be tailored to include additional or different source classes and can be applied to other X-ray catalogs. The transparency of the classification compared to other methods makes it a useful tool in the search for homogeneous populations or rare source types, including multi-messenger events. Such a tool will be increasingly valuable with the development of surveys of unprecedented size, such as LSST, SKA, and Athena, and the search for counterparts of multi-messenger events.

Classifying Unidentified X-Ray Sources in the Chandra Source Catalog Using a Multiwavelength Machine-learning Approach

“…We also crossmatch CCGCS to the TD of a recent large-scale ML-based classification study of 4XMM-DR10 (Tranin et al 2022). The recall rates calculated from the crossmatching are 93.6%, 92.0%, 36.3% for AGNs, STARs, XRBs, which are comparable to those estimated from the confident recall CM in Figure 8 while the recall rate for CVs is 25% with only 8 sources crossmatched.…”

“…The recall rates calculated from the crossmatching are 93.6%, 92.0%, 36.3% for AGNs, STARs, XRBs, which are comparable to those estimated from the confident recall CM in Figure 8 while the recall rate for CVs is 25% with only 8 sources crossmatched. We check those sources that have discrepant classifications from our results and Tranin et al (2022), and we find that most of them are from nearby, resolved galaxies or globular clusters, which are complicated and crowded environments where our MUWCLASS pipeline is not expected to work primarily due to the limitations of the MW surveys we are currently using (i.e., primarily confusion when crossmatching sources). There may also be some misclassifications in these complex environments in the TD compiled by Tranin et al (2022).…”

“…We check those sources that have discrepant classifications from our results and Tranin et al (2022), and we find that most of them are from nearby, resolved galaxies or globular clusters, which are complicated and crowded environments where our MUWCLASS pipeline is not expected to work primarily due to the limitations of the MW surveys we are currently using (i.e., primarily confusion when crossmatching sources). There may also be some misclassifications in these complex environments in the TD compiled by Tranin et al (2022). For instance, two sources (2CXO J031818.8-663230 and 2CXO J013647.4+154744) claimed to be XRBs by Tranin et al (2022) are classified by MUWCLASS as LM-STARs and actually do appear to be foreground stars (based on their parallaxes and/or proper motions 28 in Gaia EDR3), which are coincident by chance with two resolved, nearby galaxies.…”

Multiwavelength Search for the Origin of IceCube's Neutrinos

“…Due to the usually wide diversity of sources with different natures in an X-ray catalog such as the 2SXPS (with typically AGNs and stars as the most dominant populations), we applied an additional selection criterion that allows only AGNs to be in our sample. Tranin et al (2022) applied a robust probabilistic classification of X-ray sources in the Swift-XRT and XMM-Newton catalogs. We selected such X-ray sources, for which the posterior probability of being an AGN is >0.99 according to their probabilistic classification.…”

Multiwavelength search for the origin of IceCube's neutrinos

“…Due to the usually wide diversity of source natures in an X-ray catalog such like the 2SXPS (with typically AGN and stars as the most dominant populations), we needed to make an additional selection criteria that allows only AGN being in our sample. Tranin et al (2022) applied a robust probabilistic classification of X-ray sources to Swift-XRT and XMM-Newton catalogs, based on multiwavelength data, source class, and variability properties. We selected such X-ray sources, for which the posterior probability of being an AGN is > 0.99 according to their probabilistic classification.…”