Research map / Graph learning / GNN architectures
Bi-GCN: Binary Graph Convolutional Network
Check 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 |
|---|---|---|---|
| Binarize both parameters and node features | GCN · Cora | accuracy: 81.4 ± 0.4 -> 81.2 ± 0.8 | No clear effect |
| Binarize both parameters and node features | GCN · PubMed | accuracy: 79.0 ± 0.3 -> 78.2 ± 1.0 | No clear effect |
| Binarize feature extraction | inductiveGCN · Reddit | F1-micro: 93.8 ± 0.1 -> 93.1 ± 0.2 | Hurt |
| Binarize feature extraction | GraphSAGE · Reddit | F1-micro: 95.2 ± 0.1 -> 95.3 ± 0.1 | Helped |
| Binarize feature extraction | GraphSAINT · Reddit | F1-micro: 95.9 ± 0.1 -> 95.7 ± 0.1 | Hurt |
| Binarize feature extraction | inductive GNNs · Flickr | F1-micro: 50.9 ± 0.3 -> 50.2 ± 0.4; 50.9 ± 1.0 -> 50.2 ± 0.4; 52.1 ± 0.1 -> 50.8 ± 0.2 | Hurt |
| Binarize node features only (ablation) | GCN · Cora and PubMed | accuracy: Cora: 81.4 ± 0.4 -> 81.1 ± 0.4; PubMed: 79.0 ± 0.3 -> 79.4 ± 1.0 | No clear effect |
| Binarize parameters only (ablation) | GCN · Cora | accuracy: 81.4 ± 0.4 -> 78.3 ± 1.5 | Hurt |
| Binarize parameters only (ablation) | GCN · PubMed | accuracy: 79.0 ± 0.3 -> 75.5 ± 1.4 | Hurt |
| Binarize both parameters and features rather than parameters only (ablation) | Bi-GCN · Cora | accuracy: 78.3 ± 1.5 -> 81.2 ± 0.8 | Helped |
| Binarize both parameters and features rather than parameters only (ablation) | Bi-GCN · PubMed | accuracy: 75.5 ± 1.4 -> 78.2 ± 1.0 | Helped |
| Binarize both parameters and node features | GCN · Cora | model size: 360K -> 11.53K | Helped |
Closest papers by meaning
- Binary Graph Neural Networks (CVPR 2021 · similarity 0.92)
- L2-GCN: Layer-Wise and Learned Efficient Training of Graph Convolutional Networks (CVPR 2020 · similarity 0.92)
- Simplifying Graph Convolutional Networks (ICML 2019 · similarity 0.91)
- Deeper Insights Into Graph Convolutional Networks for Semi-Supervised Learning (AAAI 2018 · similarity 0.91)
- Learning Discrete Structures for Graph Neural Networks (ICML 2019 · similarity 0.91)
- Generalizing CNNs to graphs with learnable neighborhood quantization (NeurIPS 2024 · similarity 0.88)
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.