Data: 04.11.2025

Szymon Nakoneczny (CFT,PL)

QZO: A Catalog of 5 Million Quasars from the Zwicky Transient Facility

Streszczenie:
I will present a catalog of quasars from the Zwicky Transient Facility, obtained from processing the ZTF light curves with a transformer artificial neural network. Based on ZTF and WISE observations, we find 4,85M objects classified as QSOs with confidence higher than 90% (QZO). For 33% of QZO objects, with available WISE data, we publish redshifts with estimated error Delta(z)/(1 + z) = 0.14. During my talk, I will explain the machine learning methodology, present the QZO catalog, show a comparison of QSO classification results for ZTF, Pan-STARRS, WISE, and Gaia surveys, as well as analyse how QSO classification quality depends on light curve characteristics, such as timespan and median cadence.