Past Event: Babuška Forum

Scientific Machine Learning: Bridging Artificial Intelligence and Fundamental Sciences

Alfio Quarteroni, Politecnico di Milano and EPFL, Lausanne

Abstract

Despite significant progress in artificial intelligence, challenges such as inaccuracies, uncertainties, and limited transparency persist, often rendering it a "black box" technology. Scientific machine learning presents a compelling approach to address these issues by merging data-driven algorithms with digital models based on physical principles. This integration creates a valuable partnership between artificial intelligence and human expertise, anchored in natural laws and robust scientific methodologies. In this presentation, I will explore these ideas, focusing particularly on operator learning techniques, and demonstrate their application through a concrete example: the creation of a mathematical simulator designed to precisely model cardiac function.

Biography

Alfio Quarteroni is an Emeritus Professor at Politecnico di Milano and at EPFL, Lausanne. He is the founder of MOX at Politecnico di Milano. Quarteroni is a member of several prestigious academies, including the Accademia Nazionale dei Lincei, the European Academy of Sciences, the Academy of Europe, the Lisbon Academy of Sciences, and the Italian Academy of Engineering and Technology. He has authored 25 books and more than 400 research papers. Quarteroni has been honored with numerous awards, including the International Galileo Galilei Prize for Sciences in 2015, the ECCOMAS Euler Medal in 2022, the ICIAM Lagrange Prize in 2023, the Blaise Pascal Prize for Mathematics in 2024, the ECCOMAS Ritz-Galerkin medal in 2024, the SIAM Ralph Kleinman Prize in 2025.