Vorträge 1 bis 10 von 488 | Gesamtansicht
Slot saved for Fabian Gabel's external speaker.
Some peculiar (and not very well known) aspects of Gauss quadrature rules*
Thibaut Lunet, Université de Genève
Gauss quadrature rules are nowadays not only a powerful tool to compute integrals in many scientific applications, but also a numerical method that most people in the scientific community at least heard of at some point in there life.
Even if they are not the only tool to compute integral numerically, they provide the possibility to integrate any function multiplied by a given weight function (or measure), by estimating the integral of the product using a weighted sum of the function evaluations at given values (nodes).
Classical measures are well known (e.g Legendre, Chebyshev, Laguerre, Hermite), and their associated quadrature rules are well studied and documented in the literature.
While some measures allow to estimate integrals over infinite intervals (e.g Laguerre or Hermite), others also allow to integrate a function with singularities (e.g Chebyshev of the first, third and fourth kind).
However, the use of non-classical measures for specific applications can also be considered, and even this is not often used in the community, many algorithms exist to compute the nodes and weights of those quadrature rules.
In this talk we will give a quick overview of those algorithms, their efficiency, numerical stability, and some current challenge that still need to be solved.
Furthermore, under some conditions, all Gauss quadrature rules share some common properties, in particular when considering a large number of nodes.
We will give a quick overview of those common asymptotic properties, and show how they can be generalized to other applications (e.g barycentric Lagrange interpolation).
While some of those properties have been proven in particular cases, we will present some situations where they have not been proved theoretically yet, or still need to be verified.
|07.06.21||15:00||Zoom (same as Coffee Chat)||
On convergence rates of form-induced semigroup approximation
Solving evolution equations numerically requires discretizing both in time and in space. However, these two problems can be treated seperately. A common approach to spatial discretization relies on solving the weak formulation on finite-dimensional subspaces. On a semigroup level, this corresponds to approximating a semigroup by semigroups on finite-dimensional subspaces. For practical applications, quantifying the convergence speed is essential. This can be achieved by the quantified version of the Trotter-Kato theorem presented in this talk. Rates of strong convergence are obtained on dense subspaces under a joint condition on properties of both the form and the approximating spaces. An outlook to evolution equations with random coefficients and their polynomial chaos approximation will be given as well as a generalization allowing to treat the Dirichlet-to-Neumann operator.
Image reconstruction from scattered Radon data by weighted kernel functions
Positive definite kernel functions are powerful tools, which can be used to solve a variety of mathematical problems. One possible application of kernel-based methods is the reconstruction of images from scattered Radon data, which is described in . More precisely, the authors introduced weighted kernel functions to solve the reconstruction problem via generalized interpolation. Although the reconstruction method was quite competitive in comparison to standard Fourier-based methods, a detailed discussion on well-posedness and stability was mainly missing.
In this talk, I will explain the basics of kernel-based generalized interpolation and discuss the well-posedness of the proposed reconstruction method. Like most kernel-based methods, the reconstruction method also suffers from bad condition numbers. I will show how to apply well-known stabilization methods from standard Lagrangian interpolation to the generalized case to improve the stability significantly.
 S. De Marchi, A. Iske, G. Santin. Image reconstruction from scattered Radon data by weighted kernel functions. Calcolo 55, 2018.
Training of YOLO with altered activation function [Bachelorarbeitsvortrag]
Minh An Pham
* Vortrag im Rahmen des Kolloquiums für Angewandte Mathematik