AN INTRODUCTION TO SCIENTIFIC MACHINE LEARNING

Ph.D. Programme in Civil and Environmental Engineering, International cooperation and Mathematics (DICACIM)
(A.Y. 2025/26)

  
   Moodle page

Content of the course

Scientific Machine Learning (SciML) is a recently emerged research field that combines physics-based and data-driven models to numerically approximate differential equations. We aim at understanding the foundational concepts of machine learning and how they apply to problems in science and engineering. We describe key methodologies used in SciML, including data-driven modeling, physics-informed neural networks (PINNs), and hybrid modeling approaches.

Programme

• An introduction to mathematical and numerical methods to solve Partial Differential Equations.
• Foundations of supervised Machine Learning.
• Artificial Neural Networks: approximation and generalization error, optimization methods for training, backpropagation.
• Interaction between physics-based models and data-driven models.
• Surrogate modelling of high-fidelity digital models.
• Physics-informed learning.
• Operator Learning.

Course schedule

May 11, 2026 h: 9.00 - 11:00
May 13, 2026 h: 9.00 - 11:00
May 15, 2026 h: 9.00 - 11:00
May 18, 2026 h: 9.00 - 11:00
May 20, 2026 h: 9.00 - 11:00

The course will be delivered in person in the "Aula Seminari" of the "Sezione Matematica del DICATAM", ground floor of the red building of the Engineering Campus - UniBS. Gmaps 45.565610, 10.232701

Paola Gervasio   -   April 2026