<div dir="ltr"><div dir="ltr"><div dir="ltr"><div dir="ltr"><div dir="auto"><div dir="ltr"><div dir="ltr"><div dir="ltr"><div><font size="2" style="color:rgb(0,0,0)">Dear all,</font></div><font size="2" style="color:rgb(0,0,0)"><br></font>This is a gentile reminder about the inaugural lecture of the PAV-IA, which will be held <font size="2">in </font><font size="2" style="color:rgb(0,0,0)">Aula Beltrami in the Department of
Mathematics, University of Pavia on </font><b>January</b><font size="2" style="color:rgb(0,0,0)"><b> 14th 2025 from 14:30 to 15:30 followed by a coffee break</b>! <br></font></div><div dir="ltr"><br></div><div dir="ltr"><span style="background-color:rgb(255,255,255);color:rgb(255,0,0)"><b><u>For organizational purposes, we remind you that registration is free but mandatory<font size="2">. You can find the form at the following <a href="https://forms.gle/ZT2fTaMQo12JBc8w7" rel="noreferrer" target="_blank">link.</a></font></u></b></span><span style="background-color:rgb(255,255,255);color:rgb(255,0,0)"><b><u><font size="2"><br></font></u></b></span></div><div dir="ltr"><font size="2" style="color:rgb(0,0,0)"><br><br></font></div><div style="text-align:center"><font size="4" style="color:rgb(0,0,0)"><u><b>Lecture Details:</b></u></font></div><div style="text-align:center"><font size="2" style="color:rgb(0,0,0)"><u><b><br></b></u></font></div><div style="text-align:center"><font size="4" style="color:rgb(0,0,0)"><b> Speaker: </b><b><br></b></font></div><div style="text-align:center"><font size="2" style="color:rgb(0,0,0)"><b><br></b></font></div><div dir="ltr"><div style="text-align:center"><font size="2" style="color:rgb(0,0,0)"><a href="https://regazzoni.faculty.polimi.it/" rel="noreferrer" target="_blank">Francesco Regazzoni</a> </font><span>Politecnico di Milano</span></div><div style="text-align:center"><font size="2" style="color:rgb(0,0,0)"><b> </b></font></div><div style="text-align:center"><font size="4" style="color:rgb(0,0,0)"><b> Title: </b></font></div><div style="text-align:center"><font size="2" style="color:rgb(0,0,0)"><br></font></div><div style="text-align:center"><font size="2" style="color:rgb(0,0,0)">Integrating physics-based models with machine learning for fast and accurate simulations<br><br></font><font size="4" style="color:rgb(0,0,0)"><b> Abstract: </b><br></font></div><div style="text-align:center"><font size="2" style="color:rgb(0,0,0)"><br></font></div><div style="text-align:center"><div style="text-align:left"><div style="text-align:left"><font size="2" style="color:rgb(0,0,0)">Mathematical models based on differential equations, such as Partial
Differential Equations (PDEs) and Stochastic Differential Equations
(SDEs), can yield quantitative predictions of physical processes.
However, model development requires a deep understanding
of the physical processes, that is not always available. Furthermore,
the computational cost that accompanies the (possibly many-query)
numerical approximation of such mathematical models may be prohibitive
and hinder their use in relevant applications. In
this talk, we present scientific machine learning methods that
integrate physical knowledge with data-driven techniques to accelerate
the evaluation of differential models and address many-query problems -
such as sensitivity analysis, robust parameter estimation,
and uncertainty quantification. To speed up input-output evaluations,
we present <a href="https://arxiv.org/pdf/2204.07805" target="_blank">Universal Solution Manifold Networks</a>, namely emulators of
differential models capable of predicting spatial outputs and accounting
for the variability of the computational domain.
Our method is based on a mesh-less architecture, thus overcoming the
limitations associated with image segmentation and mesh generation
required by traditional discretization methods, and encodes geometrical
variability through an automatic shape encoding
technique. Furthermore, we present <a href="https://www.nature.com/articles/s41467-024-45323-x" target="_blank">Latent Dynamics Network</a>, a
space-time operator-learning method, which shows, tested in challenging
problems, superior accuracy (normalized error 5 times smaller) with
significantly fewer trainable parameters (more than 10
times fewer) than state-of-the-art methods. Numerical results
demonstrate that these scientific machine learning methods enhance
efficiency and accuracy in approximating quantities of interest, as well
as in solving parameter estimation and uncertainty quantification
problems.<br></font></div><font size="2" style="color:rgb(0,0,0)"><br></font></div><font size="4" style="color:rgb(0,0,0)"><b> Where and when: </b><b><br></b></font></div><div style="text-align:center"><font size="2" style="color:rgb(0,0,0)"><b><br></b></font></div><div style="text-align:center"><font size="2" style="color:rgb(0,0,0)">Tuesday, </font>January<font size="2" style="color:rgb(0,0,0)"> 14th, from 14:30 to 15:30 , at the Aula Beltrami, Department of Mathematics, University of Pavia.</font></div><div style="text-align:center"><font size="2" style="color:rgb(0,0,0)"><br></font></div><div style="text-align:center"><font size="2" style="color:rgb(0,0,0)"><br></font></div><font size="2" style="color:rgb(0,0,0)"><br></font></div><div dir="ltr"><font size="2" style="color:rgb(0,0,0)">We look forward to your participation!<br><br>Stay updated on SIAM Chapter activities by visiting <a href="https://sites.google.com/view/siam-unipv/home/" rel="noreferrer" target="_blank">our website</a>.<br><br>Sincerely, <br><br>PAV-IA Organizing Committee</font></div></div></div></div></div>
</div>
</div>
</div>