Fermi superluminal sources

(2) In terms of correlations: 1) both sources show positive, mutually correlated fluxes, which become closer in de-beamed fluxes; 2) with respect to apparent velocities and {gamma}-ray luminosity, there is a tendency for the brighter sources to have higher velocities; 3) with regard to viewing angle and observed {gamma}-ray luminosity, log{phi}=-(0.23+/-0.04)logL{gamma}+(11.14+/-1.93), while for the co-moving viewing angle and the intrinsic {gamma}-ray luminosity, log{phi}_co_=(0.09+/-0.01)logL^in^_(gamma)_-(1.73+/-0.48).

These correlations show that the luminous {gamma}-ray sources have smaller viewing angles and a larger co-moving viewing angle, which indicate a stronger beaming effect in {gamma}-ray emissions. Paper II (2020AN....341..462X): In our previous work in Xiao et al. (SCPMA, 2019, 62, 129811), we suggested that six superluminal sources could be {gamma}-ray candidates, and in fact, five of them have been confirmed in the fourth Fermi-LAT source catalog (4FGL).

In this work, based on the 4FGL, we report a sample of 229 Fermi detected superluminal sources (FDSs), including 40 new FDSs and 62 non-FDSs. Thus, we believe that all superluminal sources should have {gamma}-ray emissions, and superluminal motion could also be a clue to detect {gamma}-ray emission from active galactic nuclei. We present a new approach of Doppler factor estimate through the study of the {gamma}-ray luminosity (L{gamma}) and of the viewing angle ({phi}).

Paper III (2020A&C....3200387X): In this work, Machine Learning (ML) methods are used to efficiently identify the unassociated sources and the Blazar Candidate of Uncertain types (BCUs) in the Fermi-LAT Third Source Catalog (3FGL). The aims are twofold: (1) to distinguish the Active Galactic Nuclei (AGNs) from others (non-AGNs) in the unassociated sources; (2) to identify BCUs into BL Lacertae objects (BL Lacs) or Flat Spectrum Radio Quasars (FSRQs).

Two dimensional reduction methods are presented to decrease computational complexity, where Random Forest (RF), Multilayer Perceptron (MLP) and Generative Adversarial Nets (GAN) are trained as individual models. In order to achieve better performance, the ensemble technique is further explored. It is also demonstrated that grid search method is of help to choose the hyperparameters of models and decide the final predictor, by which we have identified 748 AGNs out of 1010 unassociated sources, with an accuracy of 97.04%.

Within the 573 BCUs, 326 have been identified as BL Lacs and 247 as FSRQs, with an accuracy of 92.13%.