Freitag, 22. Oktober 2021, 12:00 - 13:00 iCal

Tensor approximation in differential equations

DSUniVie Talk with Vladimir Kazeev, Department of Mathematics: Tensor approximation for discovering and exploiting structure in differential equations

Seminarraum 7, OG01
Kolingasse 14-16, 1090 Vienna

Hybrider Event (an einem physischen Ort und online)


DSUniVie Talks: The lectures on "What is Data Science @ Uni Vienna" will focus on researchers from the University of Vienna who are involved in data science or apply its methods in their field of research. The range of topics covered in these talks is as broad as the areas of research at the University of Vienna, ranging from humanities and cultural studies to chemistry, geosciences, computer science, law, history, economics and many more.

 

Our first Talk will take place on 22 October 2021 @ 12:00 CEST with Vladimir Kazeev (Department of Mathematics) about Tensor approximation for discovering and exploiting structure in differential equations.

 

ON-SITE

University of Vienna, Kolingasse 14-16, 1090 Vienna

@ Seminarraum 7, OG01

 

Registration requirements:

mailto info.datascience@univie.ac.at until 21 October at the latest (17 persons room capacity)

 

Participation requirements:

access test will be checked at the entrance of Seminarraum 7 (accepted types of evidence: vaccinated, tested or recovered)

 

ONLINE

zoom.us/j/94292120386

Meeting-ID 942 9212 0386

Passcode DStalks21

 

Abstract

Data sciences and scientific computing alike heavily rely on the approximation of large-scale data in suitable low-dimensional subspaces that reveal the most significant features or parameters and thereby dramatically reduce the complexity of the data. Mathematical models represented by differential equations are solved numerically in the form of discretizations, which essentially are sequences of subspaces of increasing dimension and of improving approximation power that are enumerated by a discretization parameter. The choice of the discretization parameter is a tradeoff between the complexity and the accuracy of the numerical method. For particular classes of differential equations, specialized methods provide more efficient sequences of approximation subspaces, designed analytically (by hand), in order to improve the approximation power without increasing the complexity. In this talk, we will discuss how tensor-network approximation, based on the separation of variables and developed originally for the simulation of many-body systems, can be leveraged to construct efficient discretization subspaces adaptively in the course of computation and to extract and exploit the multilevel structure hidden in partial differential equations.


Veranstalter

Data Science @ Uni Vienna


Um Anmeldung wird gebeten


Kontakt

Petra Schönfelder
Research Network Data Science @ Uni Vienna
+43 1 4277 50586
petra.schoenfelder@univie.ac.at