287,872 Supermassive Black Holes Masses: Deep Learning Approaching Reverberation Mapping Accuracy

Abstract

We present a population-scale catalogue of 287,872 supermassive black hole masses with high accuracy. Using a deep encoder-decoder network trained on optical spectra with reverberation-mapping (RM) based labels of 849 quasars and applied to all SDSS quasars up to $z=4$, our method achieves a root-mean-square error of $0.058$\,dex, a relative uncertainty of $\approx 14\%$, and coefficient of determination $R^2\approx 0.91$ with respect to RM-based masses, far surpassing traditional single-line virial estimators. Notably, the high accuracy is maintained for both low ($<10^{7.5}{M}_{\odot }$) and high ($>10^9{M}_{\odot }$) mass quasars, where empirical relations are unreliable.

14 pages, 9 figures. Submitted to Journal of High Energy Astrophysics