What can machine learning be used for in domain decomposition methods?

Axel Klawonn
University of Cologne
Thursday 23rd March 2023
Aula Saleri
Domain decomposition methods are a well-known approach for the efficient solution of discretized partial differential equations on parallel computers. Scientific machine learning (SciML), where techniques from machine learning and scientific computing are combined, has become an important research area in the intersection of mathematics and computer science. In this talk, we will consider a very specific field of SciML, the combination of machine learning and domain decomposition methods. Our main focus is on the use of machine learning algorithms in order to improve existing domain decomposition methods. If time allows for, we will also consider the use of domain decomposition methods in machine learning algorithms. Contatto: paola.antonietti@polimi.it This seminar is organized within the PRIN 2017 Research project «Virtual Element Methods: Analysis and Applications» Prot. 201744KLJL, funded by MIUR, Project coordinator Prof. Paola F. Antonietti