Proseminar, bachelor, winter 2024/25

Content

Making predictions on graphs, e.g., classifying nodes, classifying edges, and predicting missing edges, is important for many applications. For example, to identify spammers in social networks, to predict the type of nodes in knowledge graphs, to develop novel drugs against diseases, and to predict their properties without synthesizing them. This seminar explores recent trends in the area of graph machine learning, including feature engineering for graphs, rule learning for graphs, graph neural networks, and reasoning on graphs.

The students will read state-of-the art research papers in the area of graph machine learning and present their findings both in an oral presentation as well as in a written report.

The seminar covers:

• Feature engineering for graphs
• Rule learning for graphs
• Graph neural networks
• Reasoning on graphs

Related Work

• Stanford Lecture "Machine Learning with Graphs".
  Slides: https://web.stanford.edu/class/cs224w/
  Videos: https://www.youtube.com/watch?v=JAB_plj2rbA&ab_channel=StanfordOnline