Classification of jets in active galactic nuclei using machine learning

This work explores the application of machine learning techniques to classifying jetted active galactic nuclei (AGN) based on Very-Long-Baseline Interferometry (VLBI) observations at frequencies 1-90 GHz. Building upon previous work by Fanaroff and Riley, who classified relativistic jets in radio galaxies on kiloparsec scales, we extend this classification to parsec scales, closer to the central supermassive black hole. This approach enables detailed study of jet spatial structures and can help enhancing accuracy in global positioning systems. We define four morphological classes: single Gaussian source, double Gaussian source, and sources with single or double-sided jets. Synthetic models of AGN jets were generated to create a training dataset for a convolutional neural network (CNN). The CNN was trained on these synthetic data and subsequently applied to classify 130 thousand AGN jet images from the Astrogeo database. The distribution of images into designated classes, predicted by CNN, qualitatively matches the expected outcome.