Surrogate models and scientific machine learning at the center of a new series of events
The initiative is organized by a Spoke 9 young researcher
Trieste, 11th December 2025
Four events featuring lectures, seminars, and scientific collaborations with experts from Italian and international universities: this is the initiative organized by Federico Pichi, a young researcher from Spoke 9, as part of the project Surrogate Models for Scientific Machine Learning, funded by iNEST through a Young Researcher Grant.
The series aims to promote the exchange of advanced expertise among early-career scholars. The core of the initiative will focus on the study and development of surrogate models, mathematical tools that provide simplified representations of complex phenomena, significantly reducing computational costs compared to traditional simulations. The events will also explore the application of surrogate models to scientific machine learning, the area of artificial intelligence that integrates machine learning and scientific computing to accelerate the simulation, prediction, and analysis of complex physical systems.
Program
11th December 2025, 2:00 p.m.
Deep symmetric autoencoders from the Eckart-Young-Schmidt perspective
Nicola Rares Franco (Politecnico di Milano)
15th December 2025, 11:00 a.m.
Autoencoder embeddings: retrieving useful information from a latent space
Ettore Saetta (Università degli Studi di Napoli Federico II)
16th December 2025, 3:30 p.m.
Predicting spatiotemporal chaos: Latent vs manifold vs quantized reduced-order models
Luca Magri (Imperial College London)
16th December 2025, 4:30 p.m.
Real-time digital twins: Integrating ensemble data assimilation, reduced-order modelling, and reinforcement learning
Andrea Novoa (Imperial College London)
