Classification of Chandra X-ray Sources in Cygnus OB2

Abstract

We have devised a predominantly Naive Bayes method to classify the optical/IR matches to X-ray sources detected by Chandra in the Cygnus OB2 association into foreground, member, and background objects. We employ a variety of X-ray, optical, and infrared characteristics to construct likelihoods using training sets defined by well-measured sources. Combinations of optical photometry from SDSS (riz) and IPHAS (riHa), IR magnitudes from UKIDSS and 2MASS (JHK), X-ray quantiles and hardness ratios, and estimates of extinction Av are used to compute the relative probabilities that a given source belongs to one of the classes. We use Principal Component Analysis of photometric magnitude combinations to isolate the best axes for classification. We incorporate measurement errors into the classification. We evaluate the accuracy of the classification by inspection and reclassify a number of sources based on IR magnitudes, presence of disks, and X-ray spectral hardness. We also consider systematic errors due to extinction. We find that about 6100 objects are association members, 1400 are background, and 500 are foreground objects. The overall classification accuracy is 95%.