Miloš Kovacecvic, Graziano Chiaro, Gino Tosti, Sara Cutini.

 

In the Fermi-LAT Fourth Source Catalogue (3FGL) about 50% of the sources have no clear association with a likely γ-ray emitter. We use machine learning techinque aimed at distinguishing BL Lacs from FSRQs to investigate the source subclass of uncertain (BCU) or  unassociated ( UCS) sources characterised by γ-ray properties very similar to those of Active Galactic Nuclei.

This work is a follow up of previous papers : https://arxiv.org/abs/1607.07822 , https://arxiv.org/abs/1705.09832https://arxiv.org/abs/1808.05881https://arxiv.org/abs/1602.00385  and will use the 2019 optimization of the original algorithm as described in : Optimizing neural network techniques in classifyingFermi-LAT-ray sources.

The result of this study will suggest a new zoo for 4FGL γ-ray objects, opening up new considerations on the population of the γ-ray sky, and it will facilitating the planning of significant samples for rigorous analyses and multiwavelength observational campaigns.

 

4FGL BCUs Classification

Class        1FGL           2FGL          3FGL             4FGL

BL Lac    295 (44%)    436 (41%)    660 (38%)     1116 (36%)

FSRQ      278 (42%)    370 (35%)    484 (28%)     686(22%)

BCU           92 (14%)    257 (24%)    573 (34%)    1329(42%)

Total            665            1063              1717              3131

Table 1. Blazar class distribution in Fermi-LAT catalogs.

 

Classifying  BCUs, using a supervised machine learning method based on an artificial neural  network, probabilities for each  of 1329 uncertain blazars to be a BL Lac or FSRQ are obtained.

Using 90% precision metric, 801 can be classified as BL Lacs and 406 as FSRQs while 122 still remain unclassified.

Here the full list  4FGL BCUs_ANN_table.txt

A spectrometric optical observation campaign will be organized to confirm the data resulting from the neural algorithm. 

  • No labels