Lehre | Angewandte Funktionalanalysis | Fakultät für Mathematik | TU Chemnitz


Springe zum Hauptinhalt
Dr. Lutz Kämmerer

Lehrveranstaltung WS 20/21

Monte Carlo and quasi-Monte Carlo methods

Tue 13:45-15:15
We focus on the numerical evaluation of integrals based on sampling values. In particular for less regular functions, a Monte Carlo method, which is just a equally weighted summation of function values taking at randomly chosen sampling nodes, yields good approximations of the integral over the considered function. This fact also holds when considering functions depending on several variables.
However, when assuming higher order regularity of the integrand deterministically chosen sampling nodes may lead to even better approximations - even when applyinng the simple idea of equally weighting the sampling values in order to approximate the integral.
During the seminar, the participants present several quasi-Monte Carlo approaches and discuss advantages and disadvantages also in comparison to Monte Carlo.
The seminar will be a web-based course using BBB: https://webroom.hrz.tu-chemnitz.de/gl/lut-wuh-9aq
The next meeting takes place on Tue Oct 20th, 2020, 1:45pm.
Interested students are still more than welcome.
Contribution of each participant: 90 minutes talk and written report of about 15 pages.
Support will be provided via four-eye online meeting.