Classification of Fermi-LAT blazars with Bayesian neural networks

TL;DR

This paper introduces Bayesian neural networks for classifying Fermi-LAT blazars, offering reliable uncertainty estimates and improved robustness over traditional methods, especially with small, imbalanced datasets.

Contribution

It demonstrates the effectiveness of Bayesian neural networks in classifying gamma-ray sources and provides insights into their advantages over conventional neural networks.

Findings

Bayesian neural networks yield reliable uncertainty estimates.

They outperform conventional networks on small, imbalanced datasets.

Classification results aid in population studies and observational planning.

Abstract

The use of Bayesian neural networks is a novel approach for the classification of gamma-ray sources. We focus on the classification of Fermi-LAT blazar candidates, which can be divided into BL Lacertae objects and Flat Spectrum Radio Quasars. In contrast to conventional dense networks, Bayesian neural networks provide a reliable estimate of the uncertainty of the network predictions. We explore the correspondence between conventional and Bayesian neural networks and the effect of data augmentation. We find that Bayesian neural networks provide a robust classifier with reliable uncertainty estimates and are particularly well suited for classification problems that are based on comparatively small and imbalanced data sets. The results of our blazar candidate classification are valuable input for population studies aimed at constraining the blazar luminosity function and to guide future observational campaigns.