The Application of Machine Learning to Quasar and Seyfert Classification

Abstract

Abstract Machine learning can be utilized to classify spectra flagged as Active Galactic Nuclei (AGNs) belonging to Seyferts or Quasars, expediting data collection and aiding in analyzing the AGN types. While many properties of Seyferts and Quasars can be used as feature points in training a machine learning model, one relatively available property with high information density is the spectra of the AGN types. This paper aims to describe the training and results of a K-Nearest Neighbors and a Dense Neural Network machine learning model built to classify AGNs as Seyfert type 1s, Seyfert type 2s, or Quasars.