Automated Method Enhances Identification of Distant Quasars

Researchers have developed an automated method to identify the most distant quasars, which are crucial for understanding the early universe. The method combines Bayesian model comparison and a likelihood-based goodness-of-fit test, resulting in a high-redshift quasar selection pipeline. This pipeline was tested on data from the Sloan Digital Sky Survey (SDSS) and the UKIRT Infrared Deep Sky Survey (UKIDSS), achieving an AUC score of up to 0.795 and an F3 score of up to 0.79. The authors, Lena Lenz, Daniel J. Mortlock, Boris Leistedt, Rhys Barnett, and Paul C. Hewett, suggest that this method can be applied to upcoming surveys like Euclid, LSST, and Roman, potentially identifying hundreds of high-redshift quasars. This advancement could significantly enhance our understanding of the early stages of the universe and the formation of the first galaxies.