Gamma Ray Artificial Intelligence System (GRAIS) is a project that aims to utilise Machine Learning techniques for the automatic identification of short-duration, high-energy transient cosmic events, such as orphan afterglows, using the vast datasets from space missions like FERMI.

The GRAIS project has been selected and funded under the ICSC Programme, financed as part of the PNRR, funded by the European Union – NextGenerationEU.

Identifying short-duration transient events, like orphan afterglows, poses a significant challenge due to their rarity and the difficulty in observing the primary event. Existing catalogs, such as FERMI’s, still contain hidden events that require innovative discovery and analysis methods to be identified. Locating these sources, especially those at high redshift, is of great interest given the scarcity of revealed sources in the early universe.

The GRAIS project employs advanced Machine Learning techniques to analyze large volumes of data and identify hidden patterns. The approach includes using Generative AI to create synthetic datasets, semi-supervised learning, temporal clustering, anomaly detection, and advanced classification. This overcomes the scarcity of direct observational data and addresses the complexity of astronomical data, improving the identification of transient candidates.