Deep learning in searching the spectroscopic redshift of quasars

Studying the cosmological sources at their cosmological rest frames is crucial to track the cosmic history and properties of compact objects. In view of the increasing data volume of existing and upcoming telescopes/detectors, we here construct a 1D convolutional neural network (CNN) with a residual neural network (ResNet) structure to estimate the redshift of quasars in the Sloan Digital Sky Survey IV (SDSS-IV) catalogue from the Data Release 16 Quasar-only (DR16Q) of the extended Baryon Oscillation Spectroscopic Surv e y on a broad range of signal-to-noise ratios, named FNet . Owing to its 24 convolutional layers and the ResNet structure with different kernel sizes of 500, 200, and 15, FNet is able to disco v er the local and global patterns in the whole sample of spectra by a self-learning procedure. It reaches the accuracy of 97.0 per cent for the velocity difference for redshift, | δv| < 6000 km s-1 , and 98.0 per cent for | δv| < 12 000 km s-1 , while QuasarNET , which is a standard CNN adopted in the SDSS routine and is constructed of four convolutional layers (no ResNet structure), with kernel sizes of 10, to measure the redshift via identifying seven emission lines ( local patterns), fails in estimating redshift of 1 ∼ 3 per cent of visually inspected quasars in the DR16Q catalogue, and it gives 97.8 per cent for | δv| < 6000 km s-1 and 97.9 per cent for | δv| < 12000 km s-1 . Hence, FNet provides similar accuracy to QuasarNET , but it is applicable for a wider range of SDSS spectra, especially for those missing the clear emission lines exploited by QuasarNET . These properties of FNet , together with the fast predictive power of machine learning, allow FNet to be a more accurate alternative for the pipeline redshift estimator and can make it practical in the upcoming catalogues to reduce the number of spectra to visually inspect.