Research map / Graph learning / GNN architectures
Learning Convolutional Neural Networks for Graphs
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 |
|---|---|---|---|
| receptive field size k=10 vs k=5 (node attributes only) | PSCN · PCT / NCI1 / PROTEIN / D&D | accuracy | Mixed |
| adding edge-attribute receptive fields merged with node fields (k=10^E) (ablation) | PSCN k=10 · MUTAG / NCI1 / PROTEIN | accuracy: MUTAG 88.95→92.63; NCI1 76.34→78.59; PROTEIN 75.00→75.89; but PCT 62.29→60.00 | Helped |
| replace CNN with logistic regression on receptive fields (ablation) | PSCN k=10 · 6 benchmark graph datasets | accuracy | Hurt |
| use normalized node degree as continuous node attribute | PSCN k=10 · social graph datasets (IMDB-B, RE-B, RE-M10k, etc.) | accuracy | Helped |
| betweenness centrality instead of WL normalization (ablation) | PSCN · benchmark graph datasets | accuracy | No clear effect |
| removing dropout on larger social datasets | PSCN · social graph datasets | accuracy | No clear effect |
Closest papers by meaning
- Convolutional Kernel Networks for Graph-Structured Data (ICML 2020 · similarity 0.93)
- Semi-Supervised Classification with Graph Convolutional Networks (ICLR 2017 · similarity 0.92)
- Simplifying Graph Convolutional Networks (ICML 2019 · similarity 0.92)
- Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering (NeurIPS 2016 · similarity 0.92)
- All in a Row: Compressed Convolution Networks for Graphs (ICML 2023 · similarity 0.91)
- Adaptive Graph Convolutional Neural Networks (AAAI 2018 · similarity 0.91)
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.