Hello :) Today is Day 111!
A quick summary of today:
- Limitations of GNNs
- Scaling GNNs to Large Graphs
Continued with XCS224W: ML with Graphs and covered the penultimate module.
Limitations of GNNs
Covered topics:
Position-aware GNNs, Identity-aware GNNs, robustness of GNNs
Scaling GNNs to Large Graphs
Covered topics:
Why scaling is non-trivial, GraphSAGE and neighbour sampling, issues with neighbour sampling, subgraph sampling, ClusterGCN
That is all for today!
See you tomorrow :)