Classification of the \emph{Fermi}-LAT Blazar Candidates of Uncertain type using eXtreme Gradient Boosting

Abstract

Machine learning based approaches are emerging as very powerful tools for many applications including source classification in astrophysics research due to the availability of huge high quality data from different surveys in observational astronomy. The Large Area Telescope on board \emph{Fermi} satellite (\emph{Fermi}-LAT) has discovered more than 6500 high energy gamma-ray sources in the sky from its survey over a decade. A significant fraction of sources observed by the \emph{Fermi}-LAT either remains unassociated or has been identified as \emph{Blazar Candidates of Uncertain type} (BCUs). We explore the potential of eXtreme Gradient Boosting (XGBoost)- a supervised machine learning algorithm to identify the blazar subclasses among a sample of 112 BCUs of the 4FGL catalog whose X-ray counterparts are available within 95$\%$ uncertainty regions of the \emph{Fermi}-LAT observations. We have used information from the multi-wavelength observations in IR, optical, UV, X-ray and $\gamma$-ray wavebands along with the redshift measurements reported in the literature for classification. Among the 112 uncertain type blazars, 62 are classified as BL Lacertae objects (BL Lacs) and 6 have been classified as Flat Spectrum Radio Quasars (FSRQs). This indicates a significant improvement with respect to the multi-perceptron neural network based classification reported in the literature. Our study suggests that the gamma-ray spectral index, and IR color indices are the most important features for identifying the blazar subclasses using the \emph{XGBoost} classifier. We also explore the importance of redshift in the classification BCU candidates.

16 Pages, 10 Figures, Accepted for Publication in MNRAS

References in corpus (23)

• Incremental Fermi Large Area Telescope Fourth Source Catalog
• Machine Learning in Astronomy: a practical overview
• The Fourth Catalog of Active Galactic Nuclei Detected by the Fermi Large Area Telescope -- Data Release 3
• Fermi Large Area Telescope Performance After 10 Years Of Operation
• Blazar Flaring Patterns (B-FlaP): Classifying Blazar Candidates of Uncertain type in the third Fermi-LAT catalog by Artificial Neural Networks
• Gamma-Ray Active Galactic Nucleus Type through Machine-Learning Algorithms
• The Blazar sequence and its Physical Understanding
• Supervised detection of anomalous light-curves in massive astronomical catalogs
• 3FGLzoo. Classifying 3FGL Unassociated Fermi-LAT Gamma-ray Sources by Artificial Neural Networks
• Optical spectroscopic observations of gamma-ray blazar candidates VIII: the 2016-2017 follow up campaign carried out at SPM, NOT, KPNO and SOAR telescopes
• Classification of New X-ray Counterparts for Fermi Unassociated Gamma Ray Sources Using the Swift X-Ray Telescope
• Multiwavelength Spectral Analysis and Neural Network Classification of Counterparts to 4FGL Unassociated Sources
• Identification of BASS DR3 Sources as Stars, Galaxies and Quasars by XGBoost
• Blazars at Very High Energies: Emission Modelling
• Characterization of Variability in Blazar Light curves
• Evaluating the classification of Fermi BCUs from the 4FGL Catalog Using Machine Learning
• Gradient boosting decision trees classification of blazars of uncertain type in the fourth Fermi-LAT catalog
• Extremely High energy peaked BL Lac nature of the TeV blazar Mrk 501
• Using Neural Networks to Differentiate Newly Discovered BL Lacs and FSRQs among the 4FGL Unassociated Sources Employing Gamma-ray, X-ray, UV/Optical and IR Data
• Classification of blazar candidates of unknown type in Fermi 4LAC by unanimous voting from multiple Machine Learning Algorithms
• Classifying blazar candidates from the 3FGL unassociated catalog into BL Lacs and FSRQs using Swift and WISE data
• Artificial Neural Networks for cosmic gamma-ray propagation in the Universe
• An Artificial Intelligence based approach for constraining the redshift of blazars using $\gamma$--ray observation

Cited by in corpus (5)

• Transferring spectroscopic stellar labels to 217 million Gaia DR3 XP stars with SHBoost
• Central Engine and Spectral Energy Distribution Properties of High Redshift Gamma Ray Blazars
• kSZ Pairwise Velocity Reconstruction with Machine Learning
• Cluster optical depth and pairwise velocity estimation using machine learning
• Projected sensitivity of CTAO to axion-like particles from blazars with a machine learning approach