Distinguishing between active galactic nuclei (AGN) and star forming galaxies (SFG) using spectroscopic analysis has been done since the 1980s using a BPT diagram, using the traditional log(OIII/HB) vs log(NII/HA) lines. Through this paper, I aim to supplement the traditional emission lines used with infrared and ultraviolet photometry to find the best predictors of SFGs vs. AGNs. Successfully distinguishing between AGN and SFGs can inform us about the true demographics of these systems and how they evolve through cosmic time. I use various baseline models to predict SFG vs. AGN. For the traditional emission lines, I achieved an accuracy of 90.0%. However, with multiwavelength data, I was able to achieve an accuracy of 95.15%, indicating that there are better predictors than the ones traditionally used.
Kulkarni et al. (Sat,) studied this question.