Bachelor/Master Seminar:
Optimal Transport - from analysis to numerical methods
WS 2023/24
| Lecturer: |
Prof. Dr. Martin Huesmann Prof. Dr. Benedikt Wirth |
Information on the seminar
| Time, location: |
to be determined |
| Content: | For a few decades mathematicians have now been interested in the field of optimal transport, which has manifold applications, e.g.\ in data science, phyics, image processing, or logistics. While it was originally developed for applications, by now it also became a mathematical tool that connects several mathematical disciplines: PDE analysis, optimization, stochastics, numerics, and geometry increasingly use and extend the concept. The basic question is how some amount of material can be transported from (multiple) sources to (multiple) sinks at the lowest possible transport costs. The topic particularly gained importance due to new interpretations of statistical learning or partial differential equations via optimal transport and due to novel algorithms that allow an efficient numerical approximation of optimal transport. |
| Prerequisites: | Analysis I-III; a specialization in a module on one of the fields numerics, analysis, probability theory, geometry will be helpful. |
| Organizational meeting: |
We., 05.07.2023, 14:00-15:00, Orléansring 12, room 120.029/030 (seminar room Applied Mathematics). In case you missed that meeting, but are interested nevertheless, please let us know by e-mail. |
| Learnweb: | Learnweb-course SeminarHuesmannWirth-2023_2 |
| Requirements: | Presentation of 60-90 minutes and up to 10 pages handout for your peers (it should be shown to us and discussed with us at least a week before the presentation) |
| Topics: |
We will discuss chapters from textbooks (Peyré, Cuturi: Computational Optimal Transport) as well as research articles on those topics.
Here you find an exemplary list of topics:
|