Fermi LAT Catalog: 567 Classified Gamma-Ray Blazars via Neural Network

A catalog of 567 classified blazar candidates of uncertain type detected by the Fermi Large Area Telescope. The classifications were determined by an optimized Artificial Neural Network method, increasing performance by about 80% over a previous method and leaving only 15 sources unclassified from an original pool of 573. The table, ingested by NASA HEASARC in May 2023, orders sources by increasing likelihood of being a BL Lac type.

Use Cases

• Training classification models based on gamma-ray flux features aggregated from five energy bands.
• Analyzing the correlation between radio and X-ray flux data for blazar characterization.
• Benchmarking new machine learning methods against the reported 80% performance improvement.
• Studying the population of BL Lac objects versus other blazar types based on predicted likelihood.

Strengths

• Classification performance improved by about 80% over the previous method used in Chiaro et al. 2016.
• Covers 567 out of 573 original uncertain blazar candidates, leaving only 15 unclassified.
• Incorporates multi-wavelength data, including gamma-ray, radio (1.4 and 0.8 GHz), and X-ray (0.1-2.4 keV) fluxes.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

NASA HEASARC, based on CDS Catalog J/MNRAS/490/4770.

Collection Method

Classifications generated by an optimized Artificial Neural Network using data from the Fermi LAT 3FGL and 3LAC catalogs.

Time Range

Based on the Fermi LAT 4-Year Source Catalog and related multi-year catalogs referenced.

Freshness

Last updated 2026-03-13 17:21:38.995949; freshness should be verified.

Geography

null