Making Sense of Graphs with Machine Learning

Course Description

This course focuses on techniques for machine learning and data mining on graph data. The course starts quickly refreshing the basics of graph theory and then explores more advanced techniques of graph theory – centrality metrics (like betweenness centrality and page-rank) and community detection. The major part of the course will focus on machine learning techniques for graph and node classification. Lastly, the course will touch on the interpretability of graph machine learning models. In the practical part, we will get our hands on the selected techniques and try them with publicly available data.

Course tools

  • Python (PyTorch, pytorch_geometric, dgl, pandas, numpy libraries)
  • Jupyter
  • Google Collab

Prerequisites

  • Machine Learning Concepts (loss, gradient descent optimization, back-propagation algorithm),
  • Graph Theory Basics,
  • Python (PyTorch)
Level of complexity of course Intermediate

Lecturer

Miroslav Čepek Assistant Professor, Faculty of Information Technology, Czech Technical University Miroslav Cepek is presently an assistant professor at Czech Technical University. He is teaching advanced machine learning courses. He is also working on research projects with the university’s industrial partners. Before joining CTU, Miroslav worked for Oracle Labs on applying graph machine learning in the anti-money laundering domain. Before Oracle, Miroslav worked for Vendavo on pricing solutions for major global companies. Miroslav got his Ph.D. in Machine Learning at Czech Technical University. Fields of interests: Applied Machine Learning, Real World Applications Contacts: [email protected] https://www.linkedin.com/in/miroslav-cepek-9746246/