Explores adaptive rewiring and sparsification for scalable GNNs using Erdős-Rényi models. Tested on power grid N-1 analysis with GCN/GIN. Balances sparsity for generalization via tuning and early stopping.
Key Points
- 1.Sparsification as regularization for large graphs
- 2.Adaptive connectivity during training
- 3.Applied to electrical grid reliability
Impact Analysis
Improves GNN efficiency for real-world large-scale apps like grids. Enhances scalability without losing performance. Guides sparsity tuning.
Technical Details
Combines network science with ML on three datasets. Varies sparsity levels on GCN/GIN. Adaptive rewiring with early stopping shows promise.