Statistical analysis of cross-correlation sample of 3XMM-DR4 with SDSS-DR10 and UKIDSS-DR9

yanxia zhang,yongheng zhao,xuebing wu,haijun tian
DOI: https://doi.org/10.1017/S1743921315000022
2015-01-01
Abstract:We match the XMM-Newton 3XMMi-DR4 catalog with the Sloan Digital Sky Survey (SDSS) Data Release 10 and the United Kingdom Infrared Deep Sky Survey (UKIDSS) Data Release 9. Based on this X-ray/optical/infrared catalog, we probe the distribution of various types of X-ray emitters in the multidimensional parameter space. It is found that quasars, galaxies and stars have some kind distribution rule, especially for stars. The result shows that only using the X-ray/optical features, stars are difficult to discriminate from galaxies and quasars, the added information from infrared band is very helpful to improve the classification result of any classifier. Comparing the classification accuracy of random forests with that of rotation forests, rotation forests show better performance.
What problem does this paper attempt to address?