Our project is inspired by the increasing development and interest in scientific machine learning applied to multi-fidelity data assimilation and inverse problems. Methods to approximate probability measures have flourished recently also due to the implementation of efficient generative models in machine learning and more principled mathematical methods like diffusion models and transport maps.
KEYWORDS: machine learning, generative models, data assimilation, inverse problems
The aim of this project is to foster a new scientific collaboration between the group of Prof. Baglietto at MIT and the SISSA mathLab, a research group in mathematical modeling and scientific computing in applied mathematics. The two research teams have extremely complementary skills and expertise. The two groups share the common objective of deploying cost effective high-fidelity simulation methods that can provide increased physical accuracy and augmented engineering reliability.
dealiiX is a pioneering project aimed at developing a scalable, high-performance computational platform using the deal.II library to create accurate digital twins of human organs, with emphasis on the human brain.
This framework will leverage exascale computing capabilities and existing lighthouse applications to simulate complex biological processes in real-time, aiding in personalised medicine and advancing the diagnosis and treatment strategies of neurological disorders.
La sicurezza delle gallerie autostradali richiede strumenti capaci di supportare in modo rapido e affidabile le decisioni legate al monitoraggio e alla valutazione del rischio. Questa tesi, sviluppata nell'ambito del Master in High Performance Computing in collaborazione tra SISSA MathLab e Lombardi Group, propone uno studio preliminare sull'utilizzo di metodi numerici e tecniche di intelligenza artificiale per analizzare l'evoluzione dei difetti nelle gallerie e automatizzare la stima della relativa Classe di Attenzione.
The human cardiovascular system is a complex network of vessels, chambers, valves, and electrical signals that circulates oxygenated blood throughout the body. This complexity makes it vulnerable to diseases such as coronary artery disease, heart failure, arrhythmias, and hypertension—leading causes of global morbidity and mortality. Although mathematical and computational models can provide valuable insights, current approaches are often limited by high computational costs, excessive complexity, and poor patient-specific adaptability, restricting their clinical use.
The ROSA – Reduced Order and Surrogate Methods for Advanced Applications project, which will be developed at SISSA mathLab of the Scuola Internazionale Superiore di Studi Avanzati (SISSA), aims to advance surrogate modeling and reduced order methodologies to enable fast, reliable, and scalable simulations of complex physical systems.
H2SmartLab is designed to support the shift from traditional, non-renewable energy sources to renewable ones, with a special focus on hydrogen. Its goal is to create a research infrastructure that smartly integrates renewable hydrogen production, storage, and utilization with an advanced monitoring and management system. This integrated approach aims to enhance the resilience and efficiency of hydrogen-based systems.
