Bastian Rieck (ETH Zürich)
Learning Topology-Based Graph Representations
This talk will provide an overview of how to solve graph classification tasks using machine learning techniques. A particular emphasis will be given to techniques that are able to incorporate topological features. We will discuss (i) a novel layer that makes graph neural network architectures 'topology-aware,' and (ii) a simple, efficient method based on Betti curves that performs surprisingly well. This talk aims to be accessible to an audience of TDA enthusiasts; prior knowledge of ML techniques is not required.