GraphMLTutorialNLDL22

Tutorial NLDL22: Introduction to Graph Machine Learning, with Jupyter notebooks

This tutorial takes place during the conference NLDL 2022, on the 10th of January 2022.

In this tutorial we will get a glimpse of machine learning on graphs. The presentation is divided into 3 parts of 30 min. First we will go through an introduction to graphs, data on graphs and how to handle them using Python. Secondly, we will learn how to visualize networks in an interactive and colorful manner. Thirdly, we will design some graph neural networks and play with them. The tutorial is focused on the practical side and comes along with jupyter notebooks and Google Colabs files. The material for the course can be found here.

Part 1: Introduction to graphs with Python and Networkx

This part will go through the Jupyter notebooks:

• Loading and handling graph data 01_introduction_to_graphs.ipynb
• Basic graph properties 02_basic_graph_properties.ipynb.

Part 2: Graph visualization using Gephi

The material for this part is available here. The software Gephi is open source.

Part 3: Graph Neural Networks

• Example 1: Semi-supervised node classification
• Example 2: Graph classification

Data

The protein dataset in the data folder comes from TUdataset.