AI, Data Science & Cloud Certificates from Google, IBM & Meta
The Most Addictive Python and SQL Courses
Overview
Google, IBM & Meta Certificates – 40% Off
One Coursera Plus subscription covers most Professional Certificates on Coursera.
Unlock All Certificates
This introductory course explains why graphs are useful for non-Euclidean data and how graph neural networks operate. It covers graph representation, common graph tasks, information propagation, permutation invariance and equivariance, message passing, graph convolution, and graph attention as background for later coding implementations.
Syllabus
Introduction
Why graphs
What is a graph
Common graph tasks
Representation of a graph
- How does a GNN work?
- Understanding information propagation
- Key property: Permutation Invariance
- Key property: Permutation Equivariance
- Message passing computation
- GNN Variant: Convolution
- GNN Variant: Attention
- Ending
Taught by
Aladdin Persson