Fermi Gamma-ray Space Telescope

Searching for unclassified blazars detected by Fermi LAT using artificial neural network algorithms

GRAZIANO CHIARO
(GRAZIANO CHIARO)

Abstract:

Artificial neural network algorithms and Pass-8 event- level analysis have been used to analyze gamma-ray sources detected by Fermi LAT that remain without a conclusive classification. They represent almost 40% of the 3FGL catalog and consistent with the forthcoming FL8Y. We computed for those uncertain sources the likelihood to be BL Lacs, FSRQs or Misaligned AGN (MAGN) candidates. For some of sources light curves and prospects detectability by IACTs and CTA have been studied. Although a machine learning method cannot replace other more classical techniques for gamma-ray sources classification, it may be configured as a complementary powerful approach for the preliminary and reliable identification when multiwavelength observational data are not yet available.