Home Page

About me

I am a tenure-track assistant professor at Sapienza University of Rome (Ricercatore a Tempo Determinato di Tipo B). My research interests include the design of efficient and explainable neural network models, topological and geometric deep learning, and continual learning, as long as their applications to scientific discovery, intelligent telecommunications, medicine, and environmental monitoring.

I teach Neural Networks for Data Science Applications in the M.Sc. in Data Science, co-teach Neural Networks in the M.Sc. in AI and Robotics and the M.Sc. in Computer Engineering, and Fundamental of Machine Learning in the B.Sc. in Telecommunication Engineering. I have also held PhD courses on reproducibility, explainability, and continual learning. Recently, I wrote a book on the design of neural networks.

If you are looking for research collaborations, PhD opportunities, or thesis supervision, feel free to contact me at any moment (links in the left bar). I try to answer all emails in (max) a few days.

I am an affiliate researcher at the National Institute for Nuclear Physics (INFN), a junior fellow at the Sapienza School for Advanced Studies, and a member of the National Inter-University Consortium for Telecommunications (CNIT) and the ELLIS society. I am involved in several national and European projects, as the PI of two active Sapienza grants (DESMOS and CENTS), co-PI of MUCCA (Chist-ERA project on explainable AI for scientific discovery), task leader in 6G-GOALS (SNS-JU project on goal-oriented communications) and WP leader in RELAY (a project on relational deep learning for energy management funded by the Research Council of Norway).

Between 2017 and 2021, I was involved in several no-profit machine learning activities: I co-founded and chaired the Italian Association for Machine Learning (now closed), co-organized the Rome Machine Learning & Data Science Meetup, and co-hosted the Smarter Podcast. I was also a machine learning Google Developer Expert. Apart from here, you will find me mostly on Twitter / X, where I tweet about papers and ideas that spark my interest.

If you want to learn more about what I do, the website contains (decently updated) lists of all my academic activities, courses, publications, presentations, and grants.

Current PhD students (supervised or co-supervised)

• Federico Alvetreti (Data Science), 1st year, working on modular and efficient neural networks.
• Simone Petruzzi (Data Science), 1st year, working on large language modeling (co-funded by FastWeb, co-supervised by Fabrizio Silvestri).
• Donatella Genovese (National PhD in AI), 2nd year, working on explainable neural network models (funded by Fondazione FAIR).
• Marco Montagna (Data Science), 2nd year, working on remote sensing and topological deep learning (co-funded by NHAZCA Srl, DM 118/2023, industrial supervisor: Andrea Chessa).
• Sara Capriotti (Earth Sciences), 2nd year, working on AI applied classification of ceramic thin sections (main supervisors: Prof. Silvano Mignardi, Prof. Laura Medeghini).
• Francesco Verdini (Data Science), 2nd year, working on expressive text-to-speech synthesis (co-funded by Translated, Dottorati Industriali Regione Lazio, industrial supervisor: Dr. Sébastien Bratières).
• Ionut Marian Motoi (Computer Engineering), 3rd year, working on precision agriculture (main supervisor: Thomas Ciarfuglia).
• Alessandro Baiocchi (Data Science), 3rd year, working on adversarial machine learning and scaling up training and inference (fully funded by Leonardo Labs, industrial supervisor by Dr. Alessandro Nicolosi).
• Alessio Devoto (Data Science), 3rd year, working on modular neural network models in continual learning.
• Michele Guerra (UiT The Arctic University of Norway), 4th year, working on Bayesian neural networks and reservoir computing (main supervisor: Filippo Maria Bianchi).
• Nino Gaetano Saurio (Data Science), 3rd year, working on graph representation learning.
• Lev Telyatnikov (Data Science), incoming graduation, working on latent graph imputation.
• Alessio Verdone (ICT), incoming graduation, working on spatio-temporal graph forecasting (main supervisor: Prof. Massimo Panella).
• Roberto Benedetti (Data Science, Industrial), 3rd year, (funded by Bridgestone, industrial supervisor: Dr. Valerio Bortolotto).

Additional team members

• Jary Pomponi (post-doc fellow). Working on memory-free continual learning (former PhD student co-supervised by Aurelio Uncini).

Past students (supervised or co-supervised)

• Indro Spinelli, assistant professor (Ricercatore a Tempo Determinato di Tipo A) at the Department of Informatics of Sapienza University of Rome (former PhD student on graph representation learning co-supervised by Aurelio Uncini).
• Valerio Marsocci, former PhD student in remote sensing and earth observation co-supervised by Mattia Crespi.
• Lorenzo Lastilla, former PhD student in self-supervised learning applied to digital paleography (co-supervised by Silvia Ferrara).