Radioastronomy, even more than optical astronomy, is experiencing an explosion of volumes of observational data with the development of giant interferometers (LOFAR, ALMA, NenuFAR, SKA) that produce three-dimensional datasets (RA-DEC + frequency). Faced to these daily TB-scale data (PB-scale with SKA), the traditional methods of source detection and classification reach their limits.
In parallel, machine learning (ML) methods have undergone in recent years algorithmic and material developments that bring them to a high level of maturity. It is now appropriated to apply Artificial Intelligence (IA) to the mass of current and future radio data.
MINERVA proposes to develop and federate within the Observatoire de Paris an expertise in this (Radio-)Astronomy and Machine Learning. The members of the project are interested in a variety of astrophysical phenomena producing a radio signal (Jupiter emission, Fast Radio Bursts, synchrotron continuum of galaxies, extragalactic molecular emissions, 21 cm emission of galaxies and reinionsation, etc …) . Regardless of the astrophysical object studied, these data have sufficient similarities so that the advances obtained through artificial intelligence can spread from one domain to the other.