Conceptual

Gaussian Mixture Clustering for Changing-Look Blazar Candidates

Treats changing-look blazars -- objects that transition between FSRQ and BL Lac spectral types -- as an intermediate group in the space of eight physical blazar parameters, and applies Gaussian mixture model clustering across parameter subsets of the 4FGL gamma-ray catalog to flag new candidates. Candidate subsets are ranked by adjusted Rand index against known objects, yielding a vetted list of changing-look blazar candidates for follow-up.