A Deep-learning Anomaly-detection Method to Identify Gamma-Ray Bursts in the Ratemeters of the AGILE Anticoincidence System

Abstract

Astro-rivelatore Gamma a Immagini Leggero (AGILE) is a space mission launched in 2007 to study X-ray and gamma-ray astronomy. The AGILE team developed real-time analysis pipelines to detect transient phenomena such as gamma-ray bursts (GRBs) and react to external science alerts received by other facilities. The AGILE anticoincidence system (ACS) comprises five panels surrounding the AGILE detectors to reject background-charged particles. It can also detect hard X-ray photons in the energy range 50–200 keV. The ACS data acquisition produces a time series for each panel. The time series are merged into a single multivariate time series (MTS). We present a new deep-learning model for the detection of GRBs in the ACS data using an anomaly detection technique. The model is implemented with a convolutional neural network autoencoder architecture trained in an unsupervised manner, using a data set of MTSs randomly extracted from the AGILE ACS data. The reconstruction error of the autoencoder is used as the anomaly score to classify the MTS. We calculated the associated p-value distribution, using more than 107 background-only MTSs, to define the statistical significance of the detections. We evaluate the trained model with a list of GRBs reported by the GRBWeb catalog. The results confirm the model’s capabilities to detect GRBs in the ACS data. We will implement this method in the AGILE real-time analysis pipeline.

Gamma-ray Bursts: 50 Years and Counting!

“…More recently, AGILE has had huge success in the application of modern machinelearning techniques for the identification of GRB signals in data from both the gamma-ray imaging detector (GRID) [122] and the anticoincidence system (ACS) [123], leading to the identification of 72 GRB signals with significance ≥3σ, 15 of which are not present in the second MCAL GRB catalog, as they were not identified before when using traditional methods for the data analysis.…”

New detailed characterization of the residual luminescence emitted by the GAGG:Ce scintillator crystals for the HERMES Pathfinder mission

“…By including realistic background levels and varying GRB fluxes, the CNN is better prepared to handle the complexities of realworld data and improve its performance in classifying gamma-ray sky maps with GRBs. In [197], a novel anomaly detection technique for multivariate time series (MTS) is presented. The technique involves using a CNN autoencoder to calculate the anomaly score based on the MTS reconstruction error.…”

The First AGILE Solar Flare Catalog

“…We carried out a search for GRBs in the AC-Lat4 database, by cross-correlating the AC data with the comprehensive Icecube GRBWeb catalog by P. Coppin, 11 already adopted for AGILE studies on GRBs (Ursi et al 2022b;Parmiggiani et al 2023). The cross-search ended up with 145 bursts detected by the AC-Lat4.…”