Graph-based Machine Learning

• General
• Content of the lecture
• Topics
• Requirements
• Literature

General

Supervisor:

Appointments (every summer semester, including lecture and exercise classes: 5 CPs):

• Lecture: see    
• Exercise: see    

Exam:

• Written exam
• Appointments: will be announced at the beginning of the summer semester

Lecture slides: see    
Exercise materials: see    

Content of the lecture

Teaching of fundamental concepts and methods of graph-based Machine Learning and related methods for decision support. You will learn about exemplary applications of graph-based models and understand how they work. Furthermore, they will learn to select, adapt and evaluate models for new applications and how to evaluate them.

Topics

• Introduction to Graphs, Types of Graphs, Disease Propogration/Social Networks and other Applications
• Markov Processes, Markov Chains
• Markov Random Fields
• Basics of Probability Theory, Bayes Theorem, Representation of Uncertain Information
• Independence, Decomposition, Bayes Networks
• Probabilistic Reasoning, Propagation, Naive Bayes
• Parameter Learning, Structure Learning
• Causal Networks
• Graph Clustering, Random Walks, Node2Vec
• Graph Neural Networks

Requirements

We strongly recommend that you know the foundations of

• machine learning
• deep learning
• or AI in general

in order to attend the course. You should have attended at least one other course for ML and DL in the past.
This course as well as the exercises will be in English only.

Literature

• Borgelt, C., Steinbrecher, M., and Kruse, R. (2009). Graphical Models: Representations for Learning, Reasoning and Data Mining (2nd Edition). John Wiley & Sons, Chichester, United Kingdom
• Jensen, F (1996). An Introduction to Bayesian Networks. UCL Press, London, United Kingdom
• Wu, L., Cui, P., Pei, J., and Zhao, L. (2022). Graph Neural Networks - Foundations, Frontiers, and Applications. Springer, Singapore, 2022, https://graph-neural-networks.github.io/index.html