This foundational training school on Artificial Intelligence methodologies is designed to provide theoretical and practical skills to INAF personnel from diverse scientific backgrounds.
OBJECTIVES AND CONTENT
The core curriculum will cover the following key topics:
Preprocessing of astronomical data
Supervised learning models (classification and regression)
Unsupervised learning models (clustering)
Anomaly detection
The focus will remain on the theory and practice of the methods themselves rather than on specific scientific problems. By utilizing standard examples and datasets, we aim to prevent the technical dispersion often caused by differing scientific backgrounds (e.g., avoiding the need to explain the nuances of AGN to a solar physicist, or vice versa).
Each lesson will be supported by a dedicated notebook (including relevant data), which will be made available to students in the days leading up to the school.
To promote internal INAF research activities, each day will feature short talks dedicated to the scientific challenges currently being addressed by participants and instructors.
Furthermore, the program will include a round-table discussion with representatives from the INAF LLM thematic group and a technical session on the computing infrastructure currently available within the Institute.
To ensure future reproducibility of the school (and potential rotation of the venue), instructors will be selected primarily from within INAF staff. While lesson assignments have not yet been formalized, several members of the thematic group have expressed interest. Specifically, the Astroinformatics group of Naples—leveraging the experience of Stefano Cavuoti, who taught the Astroinformatics degree course at the University of Naples Federico II for three years—and Simone Riggi (INAF Catania Astrophysical Observatory) are expected to contribute significantly.
PARTECIPANTI
The "Auditorium Nazionale" has a maximum capacity of approximately 250 seats. Participation will be limited based on the allocated budget. The budget is intended to cover 6 coffee breaks and, ideally, 3 lunches. While lunches are not strictly essential, they are highly recommended given the limited availability of restaurants in the immediate vicinity of the Observatory.
Based on an estimated cost of €12.50 (+VAT) per coffee break and €25.00 (+VAT) for a light lunch, the cost per participant is approximately €170. Assuming a cap of 100 attendees, the total budget required is approximately €17k. Should the number of registered participants be lower, the excess funds will be returned to the budget.
In the event that the number of registrations exceeds the budget capacity, a "first-come, first-served" policy will be applied to ensure a democratic selection process.
Remote participation will be available for those unable to travel. Although all lectures will be recorded and made available online, it is important to note that remote attendance is significantly less effective than in-person participation for this type of hands-on training.