Research map / Graph learning / GNN architectures
Local Augmentation for Graph Neural Networks
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| Add local augmentation to GCN | GCN · Cora, Citeseer, Pubmed | accuracy: 81.5{$\pm$0.5} -> 84.6{$\pm$0.5}; 70.3{$\pm$0.7} -> 74.7{$\pm$0.5}; 79.0{$\pm$0.5} -> 81.7{$\pm$0.7} | Helped |
| Add local augmentation to GAT | GAT · Cora, Citeseer, Pubmed | accuracy: 83.0{$\pm$0.7} -> 84.7{$\pm$0.4}; 72.5{$\pm$0.7} -> 73.7{$\pm$0.5}; 79.0{$\pm$0.3} -> 81.0$\pm$0.4 | Helped |
| Add local augmentation to G-Rand | G-Rand · Cora, Citeseer, Pubmed | accuracy: +0.3%, +0.4%, and +0.7% | Helped |
| Add local augmentation to GCN | GCN · ogbn-products | accuracy: 75.64{$\pm$0.21} -> 76.11{$\pm$0.09} | Helped |
| Add local augmentation to GCN | GCN · ogbn-proteins | ROC-AUC: 72.51{$\pm$0.35} -> 73.25{$\pm$0.51} | Helped |
| Add local augmentation to GCN | GCN · ogbn-arxiv | accuracy: 71.74{$\pm$0.29} -> 72.08{$\pm$0.14} | Helped |
| Add local augmentation to GCN | GCN · ogbg-molhiv | ROC-AUC: 76.06{$\pm$0.97} -> 76.18{$\pm$1.11} | No clear effect |
| Add local augmentation to GCN | GCN · ogbg-molpcba | AP: 20.20{$\pm$0.24} -> {20.28$\pm$0.16} | Helped |
| Add local augmentation to GIN | GIN · ogbg-molhiv and ogbg-molpcba | ROC-AUC and AP: 75.58{$\pm$1.40} -> 75.20{$\pm$}1.74; 22.66{$\pm$0.28} -> {22.38$\pm$0.24} | Hurt |
| Use concatenation architecture without generated features (ablation) | GCN · Pubmed | accuracy: 79.0 -> 79.3{$\pm$0.4}; 0.3 | Helped |
| Add generated features without consistency training (ablation) | GCN with concatenation · Pubmed | accuracy: 79.3{$\pm$0.4} -> 81.1{$\pm$0.5}; 1.8 | Helped |
| Add consistency training (ablation) | GCN with local augmentation · Pubmed | accuracy: 81.1{$\pm$0.5} -> 81.4{$\pm$0.5}; 0.3 | Helped |
Closest papers by meaning
- Data Augmentation for Graph Neural Networks (AAAI 2021 · similarity 0.93)
- Community-Invariant Graph Contrastive Learning (ICML 2024 · similarity 0.88)
- Graph Neural Network Generalization With Gaussian Mixture Model Based Augmentation (ICML 2025 · similarity 0.88)
- Locality-Aware Graph Rewiring in GNNs (ICLR 2024 · similarity 0.88)
- Edge Prompt Tuning for Graph Neural Networks (ICLR 2025 · similarity 0.87)
- Through the Dual-Prism: A Spectral Perspective on Graph Data Augmentation for Graph Classifications (AAAI 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.