Classification of BATSE, Swift, and Fermi Gamma-Ray Bursts from Prompt Emission Alone

TL;DR

This study employs machine learning algorithms, t-SNE and UMAP, to classify gamma-ray bursts from BATSE, Swift, and Fermi catalogs into short and long categories based solely on prompt emission data, revealing potential additional burst types.

Contribution

It extends previous classification methods by applying two dimensionality reduction algorithms to multiple GRB catalogs, improving classification robustness and identifying possible new burst subclasses.

Findings

Clear separation of short and long GRBs in embeddings

Some bursts cannot be robustly classified

Conflicting classifications for some bursts observed by multiple satellites

Abstract

Although it is generally assumed that there are two dominant classes of gamma-ray bursts (GRB) with different typical durations, it has been difficult to unambiguously classify GRBs as short or long from summary properties such as duration, spectral hardness, and spectral lag. Recent work used t-distributed stochastic neighborhood embedding (t-SNE), a machine learning algorithm for dimensionality reduction, to classify all Swift gamma-ray bursts as short or long. Here, the method is expanded, using two algorithms, t-SNE and UMAP, to produce embeddings that are used to provide a classification for the 1911 BATSE bursts, 1321 Swift bursts, and 2294 Fermi bursts for which both spectra and metadata are available. Although the embeddings appear to produce a clear separation of each catalog into short and long bursts, a resampling-based approach is used to show that a small fraction of bursts cannot be robustly classified. Further, 3 of the 304 bursts observed by both Swift and Fermi have robust but conflicting classifications. A likely interpretation is that in addition to the two predominant classes of GRBs, there are additional, uncommon types of bursts which may require multi-wavelength observations in order to separate from more typical short and long GRBs.