Teaching

See TUMonline
Note: Registration is only possible via the matching platform.

By the end of the seminar, students have studied a specific problem in the graph machine learning domain and presented ideas on how to solve it (experimentally or with literature). In the seminar, students will learn about scientific paper writing techniques, the peer review process, and presentation of results. Learn how to answer a research question. Learn how to read and judge scientific papers. Learn how to build prototypes of algorithms. Learn how to present scientific work. Learn scientific writing skills. Deep dive into a practical graph machine learning topic.

A graph is a data structure able to capture relations between data points, e.g. connections within a social network, citations of scientific papers or traffic networks. As graph-structured data is all around us and allows us to intuitively capture a lot of real-world problems, techniques to learn on those graphs have been developed. Like this, assessments on graph-, node- or edge-level can be made. Since traditional machine learning models cannot easily be applied to graphs, the so-called Graph Neural Networks (GNNs) have been developed. In this seminar, we will focus on different GNN architectures and optimizations as well as various fields of application. Students will be able to deep dive into a specific GNN problem and present ideas on how to solve it. The topics consist of conducting literature research and/or coding your own prototype. The preliminary meeting will take place on 01.02.2024 at 1:30pm via https://tum-conf.zoom-x.de/j/64211813958?pwd=TGsyMXMwdHVOZXlna1FDOFNrdm9YQT09

The students should have a general understanding of (graph) machine learning. The course topics are suitable for all experience levels.

Conference-like structure with writing your own report, reviewing one from another group, and presenting your findings with a presentation.