Machine Learning Based Classification of Type 1 Active Galactic Nuclei

Abstract

Aims. We aim to provide a machine learning based method to distinguish type 1 quasars and type 1 Seyfert galaxies based on photometric data, which could be applied to large all-sky surveys. The method can also be used to obtain relatively complete samples of Seyfert galaxy candidates when broad-line AGNs have been selected. Results. The results of PCA show that quasars and Seyfert galaxies are partial distinguishable. The supervised classifier based on random forest (RF) achieves good performances for type 1 quasars with a F1 score of 0.967. The performance for classifying Seyfert galaxies is worse, with a F1 score of 0.791. Increasing the classification threshold can help to derive a sample of Seyfert galaxies with high purity. The classifier are then applied to the cross-matched sample with DESI DR1. When compared with the results of spectroscopic classification, quasars also show high precision and recall. For Seyfert galaxies identified by RF classifier, there are plenty of less luminous sources who are classified as quasars by spectroscopic method. If they are regarded as Seyfert galaxies, our classifier also performs well to predict Seyfert galaxies. Both the median SEDs and the fraction of Seyfert galaxies derived by our classifier are similar with those for the training sample, while the fraction of Seyfert galaxies for spectroscopic classification are lower at all redshift. These results indicate that our classifier can recover the features of Seyfert galaxies, at least those similar to the Seyfert galaxies in the training sample.