WiSe 23/24: Methods for the analysis of graphs and networks
Katharina Baum, Pauline Hiort
Comments
The world is complex, and so is its data. Graphs (or networks) are key for the analysis of complex data and in the integration of data layers. They allow mapping and formal investigation of relationships between entities (nodes).
In the course of the seminar, we will explore different computational analysis methods of graphs and networks. We will learn about different types and properties of graphs and how to deal with them in general. In addition, application areas of graphs such as social or biological networks will be discussed.
Examples for specific topics are
- Basic properties of graphs and nodes and how to determine them: shortest paths, centralities, degree distribution, clustering coefficients
- You have to go over seven nodes - the small-world property of social interaction networks
- Clusters and communities in networks
- Application: Networks for mapping molecular regulations
- Random graphs and their application
- Working with large networks, representative subgraphs
- Modeling information and signal flow in graphs and detecting signal sources
- Nodes as numbers - graph-based embedding methods
- Graphs and machine learning: graph neural networks (GNNs)
We will meet on Wednesday, October 4, 2023 at 10:15 am (s.t.) (in a hybrid setting, please check out the Whiteboard for room and/or access link) for a first introduction and topic assignment. Please contact us early (pauline.hiort@fu-berlin.de or katharina.baum@fu-berlin.de) in case you want to participate in the proseminar and cannot attend this meeting.
closeSuggested reading
Examples for employed literature
- "Network Science" by A. L. Barbarasi (http://networksciencebook.com/): Different chapters from the book Network Science about the basics of graphs and networks
- "node2vec: Scalable Feature Learning for Networks" by A. Grover and J. Leskovec (https://doi.org/10.1145/2939672.2939754): Publication about a graph-based embedding where nodes are coded as vectors
- "A Gentle Introduction to Graph Neural Networks" by B. Sanchez-Lengeling and others (https://distill.pub/2021/gnn-intro/): Article introducing the topic of graph neural networks
16 Class schedule
Regular appointments