Artificial Neural Network Classification of 4FGL Sources

TL;DR

This paper uses ensemble artificial neural networks to classify gamma-ray sources from Fermi-LAT catalogs into pulsars, BL Lac blazars, and FSRQs, identifying promising candidates among unassociated sources for further study.

Contribution

It introduces a novel ANN-based classification method applied to Fermi-LAT gamma-ray sources, improving identification of source types and highlighting potential new objects.

Findings

Identified candidate pulsars, blazars, and quasars among unassociated sources.

Selected ten outliers as promising targets for follow-up.

Achieved approximate category balance in classifications.

Abstract

The Fermi-LAT DR1 and DR2 4FGL catalogues feature more than 5000 gamma-ray sources of which about one fourth are not associated with already known objects, and approximately one third are associated with blazars of uncertain nature. We perform a three-category classification of the 4FGL DR1 and DR2 sources independently, using an ensemble of Artificial Neural Networks (ANNs) to characterise them based on the likelihood of being a Pulsar (PSR), a BL Lac type blazar (BLL) or a Flat Spectrum Radio Quasar (FSRQ). We identify candidate PSR, BLL and FSRQ among the unassociated sources with approximate equipartition among the three categories and select ten classification outliers as potentially interesting for follow up studies.