Statistical Study of 2XMMi-DR3/SDSS-DR8 Cross-correlation Sample

TL;DR

This study cross-correlates X-ray and optical catalogs to analyze the distribution of X-ray emitters, applying machine learning for classification, and highlights challenges in distinguishing X-ray active stars from quasars and galaxies.

Contribution

It provides a large X-ray/optical catalog and evaluates the effectiveness of a random forest classifier in identifying different X-ray source types.

Findings

Quasars and galaxies occupy distinct regions in photometric space.

Classification accuracy exceeds 93% for quasars and galaxies.

X-ray active stars are harder to distinguish, with only 45.3% accuracy.

Abstract

Cross-correlating the XMM-Newton 2XMMi-DR3 catalog with the Sloan Digital Sky Survey (SDSS) Data Release 8, we obtain one of the largest X-ray/optical catalogs and explore the distribution of various classes of X-ray emitters in the multidimensional photometric parameter space. Quasars and galaxies occupy different zones while stars scatter in them. However, X-ray active stars have a certain distributing rule according to spectral types. The earlier the type of stars, the stronger X-ray emitting. X-ray active stars have a similar distribution to most of stars in the g-r versus r-i diagram. Based on the identified samples with SDSS spectral classification, a random forest algorithm for automatic classification is performed. The result shows that the classification accuracy of quasars and galaxies adds up to more than 93.0% while that of X-ray emitting stars only amounts to 45.3%. In other words, it is easy to separate quasars and galaxies, but it is difficult to discriminate X-ray active stars from quasars and galaxies. If we want to improve the accuracy of automatic classification, it is necessary to increase the number of X-ray emitting stars, since the majority of X-ray emitting sources are quasars and galaxies. The results obtained here will be used for the optical spectral survey performed by the Large sky Area Multi-Object fiber Spectroscopic Telescope (LAMOST, also named the Guo Shou Jing Telescope), which is a Chinese national scientific research facility operated by the National Astronomical Observatories, Chinese Academy of Sciences.