This repo was created to contain material used for ICAC 2022 (4th International Conference on Advancements in Computing) workshop on Graph Neural Networks: Theory and Applications.
Our workshop is divided into 3 parts
- Session 1: Introduction to fundamental theory and concepts of GNNs.
- Session 2: Tutorial 1 on one of the popular GNN variant which is Graph Convolutional Neural Network.
- Session 3: Tutorial 2 on GNN application on bitcoin. Here we use GNN to detect fraudulent transaction using a publicly available bitcoin graph dataset.
Note: The notebooks are designed to run standalone, so they contain all the necessary installation setup. We highly suggest to run these notebooks in Google colab.
| Session | Slides | Colab |
|---|---|---|
| Session 1 | Graph Neural Networks Theory and Application | ----------------- |
| Session 2 | Tutorial 1: GCN Slides |  |
| Session 3 | Tutorial 2: GNN Application | |
The following are great resources to get started on GNN
- W. L. Hamilton, Graph Representation Learning Textbook( freely available text book)
- CS224: Machine Learning with Graphs 2021, Lecture Notes
- Petar Velickovic, Theoretical Foundations of Graph Neural Networks
- PyTorch Geometric Tutorials
- Deep Graph Library Tutorials
- Aleksa Gordic, GNN research paper explanations
- Learning on Graphs Conference, GNN oriented conference
- LoGaG: Learning on Graphs and Geometry Reading Group