Deep Learning Blazar Classification based on Multi-frequency Spectral Energy Distribution Data

Abstract

Blazars are among the most studied sources in high-energy astrophysics as they form the largest fraction of extragalactic gamma-ray sources and are considered prime candidates for being the counterparts of high-energy astrophysical neutrinos. Their reliable identification amid the many faint radio sources is a crucial step for multi-messenger counterpart associations. As the astronomical community prepares for the coming of a number of new facilities able to survey the non-thermal sky at unprecedented depths, from radio to gamma-rays, machine learning techniques for fast and reliable source identification are ever more relevant. The purpose of this work was to develop a deep learning architecture to identify blazar within a population of AGN based solely on non-contemporaneous spectral energy distribution information, collected from publicly available multi-frequency catalogues. This study uses an unprecedented amount of data, with SEDs for $\approx 14,000$ sources collected with the Open Universe VOU-Blazars tool. It uses a convolutional long-short term memory neural network purposefully built for the problem of SED classification, which we describe in detail and validate. The network was able to distinguish blazars from other types of AGNs to a satisfying degree (achieving a ROC area under curve of $0.98$), even when trained on a reduced subset of the whole sample. This initial study does not attempt to classify blazars among their different sub-classes, or quantify the likelihood of any multi-frequency or multi-messenger association, but is presented as a step towards these more practically-oriented applications.

13 pages, 17 figures, updated with accepted version

References in corpus (49)

• Adam: A Method for Stochastic Optimization
• Fermi Large Area Telescope Fourth Source Catalog
• A Unifying View of the Spectral Energy Distributions of Blazars
• Neutrino emission from the direction of the blazar TXS 0506+056 prior to the IceCube-170922A alert
• The Spectral Energy Distribution of Fermi bright blazars
• SUMSS: A Wide-Field Radio Imaging Survey of the Southern Sky. II. The Source Catalogue
• Rotation-invariant convolutional neural networks for galaxy morphology prediction
• The Connection between X-ray- and Radio-Selected BL Lacertae Objects
• eROSITA Science Book: Mapping the Structure of the Energetic Universe
• The XMM-Newton serendipitous survey IX. The fourth XMM-Newton serendipitous source catalogue
• The Fourth Catalog of Active Galactic Nuclei Detected by the Fermi Large Area Telescope
• 3FHL: The Third Catalog of Hard Fermi-LAT Sources
• The 5th edition of the Roma-BZCAT. A short presentation
• The first XMM-Newton slew survey catalogue: XMMSL1
• Fast Automated Analysis of Strong Gravitational Lenses with Convolutional Neural Networks
• Science with the Cherenkov Telescope Array
• Star-galaxy Classification Using Deep Convolutional Neural Networks
• 2SXPS: An improved and expanded Swift X-ray telescope point source catalog
• CMU DeepLens: Deep Learning For Automatic Image-based Galaxy-Galaxy Strong Lens Finding
• Dissecting the region around IceCube-170922A: the blazar TXS 0506+056 as the first cosmic neutrino source
• Blazar Flares as an Origin of High-Energy Cosmic Neutrinos?
• The Half Million Quasars (HMQ) Catalogue
• The Strong Gravitational Lens Finding Challenge
• Uncertainties in Parameters Estimated with Neural Networks: Application to Strong Gravitational Lensing
• The WISE gamma-ray strip parametrization: the nature of the gamma-ray Active Galactic Nuclei of Uncertain type
• The WISE Blazar-Like Radio-Loud Sources: An All-Sky Catalog Of Candidate Gamma-Ray Blazars
• The 3HSP catalogue of Extreme & High Synchrotron Peaked Blazars
• Dissecting the regions around IceCube high-energy neutrinos: growing evidence for the blazar connection
• Unveiling the nature of the unidentified gamma-ray sources I: a new method for the association of gamma-ray blazars
• 2WHSP: A catalog of HE and VHE gamma-ray blazars and blazar candidates
• Active Galactic Nuclei under the scrutiny of CTA
• 1WHSP: an IR-based sample of $\sim$1,000 VHE $\gamma$-ray blazar candidates
• DASH: Deep Learning for the Automated Spectral Classification of Supernovae and their Hosts
• Gamma-Ray Active Galactic Nucleus Type through Machine-Learning Algorithms
• PELICAN: deeP architecturE for the LIght Curve ANalysis
• The use of convolutional neural networks for modelling large optically-selected strong galaxy-lens samples
• A Neural Network Gravitational Arc Finder based on the Mediatrix filamentation Method
• The Open Universe VOU-Blazars tool
• Open Universe for Blazars: a new generation of astronomical products based on 14 years of Swift-XRT data
• Classification of New X-ray Counterparts for Fermi Unassociated Gamma Ray Sources Using the Swift X-Ray Telescope
• Analyzing interferometric observations of strong gravitational lenses with recurrent and convolutional neural networks
• Evaluating the classification of Fermi BCUs from the 4FGL Catalog Using Machine Learning
• The strange case of the transient HBL blazar 4FGL J1544.3-0649
• Optimizing neural network techniques in classifying Fermi-LAT gamma-ray sources
• Machine Learning applied to Multifrequency Data in Astrophysics: Blazar Classification
• On the physical association of Fermi-LAT blazars with their low-energy counterparts
• Deep Learning in Wide-field Surveys: Fast Analysis of Strong Lenses in Ground-based Cosmic Experiments
• A complete sample of LSP blazars fully described in $\gamma$-rays. New $\gamma$-ray detections and associations with Fermi-LAT
• The Open Universe survey of Swift-XRT GRB fields: a complete sample of HBL blazars

Cited by in corpus (3)

• Gradient boosting decision trees classification of blazars of uncertain type in the fourth Fermi-LAT catalog
• Searching for TeV Candidates in 4LAC High-synchrotron-peaked Frequency BL Lac Objects
• Developing a Victorious Strategy to the Second Strong Gravitational Lensing Data Challenge