This lab is designed to showcase and familiarise you with the basics of learning on hypergraphs. It will encompass both the theory behind hypergraph networks and putting them into practice.
The lab is split into 4 main parts.
Part 1:
- Creating the hypergraph object
- Understand how to work with hypergraph-structure data
- Building a basic HNN for node-level prediction
Part 2:
- Comparing hypergraph methods to graph based methods
Part 3:
- Creating mini-batches for hypergraphs
- Hypergraphs for graph-level prediction
Part 4:
- Hypergraph attention networks
If you find any mistakes please raise it as an issue or get in touch directly at hypergraph.practical@gmail.com !
Likewise, please get in touch with any feedback from the tutorial via email.
Daniel