Artificial Intelligence and Mathematics 2022

Seminario di Carola-Bibiane Schönlieb per il ciclo AIM 2022

Carola-Bibiane Schönlieb sarà ospite di Italia De Feis e Flavio Lombardi per il ciclo di seminari Artificial Intelligence and Mathematics 2022. Carola è professore presso il Dipartimento di Matematica Applicata e Fisica Teorica (DAMTP) dell'Università di Cambridge e presenta un talk dal titolo "Intelligenza Artificiale e imaging matematico" con il seguente abstract:

Daniela Di Serafino: On Some Research Lines in Optimization Methods for Machine Learning

Mercoledì 18 maggio Daniela Di Serafino parteciperà con un talk al ciclo di seminari Artificial Intelligence and Mathematics 2022

Daniela Di Serafino è docente di Analisi Numerica presso il Dipartimento di Matematica e Applicazioni "R. Caccioppoli" dell'Università degli Studi di Napoli Federico II.

Gitta Kutyniok - The Modern Mathematics of Deep Learning

Despite the outstanding success of deep neural networks in real-world applications, ranging from science to public life, most of the related research is empirically driven and a comprehensive mathematical foundation is still missing. At the same time, these methods have already shown their impressive potential in mathematical research areas such as imaging sciences, inverse problems, or numerical analysis of partial differential equations, sometimes by far outperforming classical mathematical approaches for particular problem classes.


Consensus-based optimization (CBO) is a multi-agent metaheuristic derivative-free optimization method that can globally minimize nonconvex nonsmooth functions and is amenable to theoretical analysis. In fact, optimizing agents (particles) move on the optimization domain driven by a drift towards an instantaneous consensus point, which is computed as a convex combination of particle locations, weighted by the cost function according to Laplace’s principle, and it represents an approximation to a global minimizer.

Constantinos Siettos - Physics Informed Random Projection Neural Networks for the Numerical Solution of the Forward and Inverse Problems in Differential Equations

Over the last few years, machine learning has been used to solve both the forward, i.e. the numerical solution of time-depended non-linear differential equations as an alternative to classical numerical analysis methods, but also the inverse problem i.e.