20–27 Mar 2026
Wild View Resorts
Africa/Gaborone timezone

Machine Learning Classification of High-Energy GRB Counterparts using Fermi-GBM and LAT Data

27 Mar 2026, 15:00
15m
Wild View Resorts

Wild View Resorts

Plot 80 President Avenue, Kasane, Botswana
Online - Talk 9 Machine Learning & techniques Science & Engineering

Speaker

Tamador Khalil Mansoor Aldowma (University of Johannesburg)

Description

Simultaneous detections of Gamma-Ray Bursts (GRBs) by the Fermi Gamma-ray Burst Monitor (GBM) and the Fermi Large Area Telescope (LAT) instruments are crucial for understanding the full spectral evolution of relativistic jets. Addressing the delay in official catalog releases, we developed a scalable machine learning pipeline to identify coincident events in real-time. By performing a systematic cross-match of available catalog data, we generated a labeled dataset of approximately 220 coincident candidates. We extract a comprehensive set of features for classification, including temporal offsets, spectral hardness ratios, fluence/flux comparisons, peak energy ($E_{peak}$), and duration. These features serve as inputs for machine learning classification models, which are trained to distinguish between physical associations and random spatial coincidences. We report on the feature importance distributions, demonstrating that specific spectral signatures are highly predictive of high-energy emission. This framework allows for the extension of the confirmed GRB catalog beyond August 2022, providing a tool for the immediate classification of the latest bursts.

Stream Science or Engineering

Primary author

Tamador Khalil Mansoor Aldowma (University of Johannesburg)

Co-author

Soebur Razzaque (University of Johannesburg)

Presentation materials

There are no materials yet.