Research map / Graph learning / GNN architectures
Beyond Graph Convolutional Network: An Interpretable Regularizer-Centered Optimization Framework
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 |
|---|---|---|---|
| combine topological and semantic regularizers using low-rank approximation | tsGCN · Cora | accuracy: 80.6 -> 82.0; JKNet: 83.1 | Hurt |
| combine topological and semantic regularizers using low-rank approximation | tsGCN · Citeseer | accuracy: 73.1 vs 72.3 | Helped |
| combine topological and semantic regularizers using low-rank approximation | tsGCN · Pubmed | accuracy: 82.4 vs 80.6 | Helped |
| combine topological and semantic regularizers using low-rank approximation | tsGCN · ACM | accuracy: 92.8 vs 91.2 | Helped |
| combine topological and semantic regularizers using low-rank approximation | tsGCN · BlogCatalog | accuracy: 92.3 vs 87.7 | Helped |
| combine topological and semantic regularizers using low-rank approximation | tsGCN · CoraFull | accuracy: 67.9 vs 65.6 | Helped |
| combine topological and semantic regularizers using low-rank approximation | tsGCN · Flickr | accuracy: 79.1 vs 60.7 | Helped |
| combine topological and semantic regularizers using low-rank approximation | tsGCN · UAI | accuracy: 67.9 vs 62.3 | Helped |
| replace the low-rank approximation with direct matrix inversion (ablation) | tsGCN · Flickr | accuracy: 79.1 -> 53.3 | Hurt |
| replace the low-rank approximation with direct matrix inversion (ablation) | tsGCN · ACM | accuracy: 92.8 -> 85.1 | Hurt |
| add the semantic regularizer to the topological regularizer (ablation) | tsGCN · node classification across datasets | accuracy | Helped |
| add the low-rank approximation for infinite-order graph convolution (ablation) | tsGCN · node classification across datasets | accuracy | Helped |
Closest papers by meaning
- Simplifying Graph Convolutional Networks (ICML 2019 · similarity 0.92)
- All in a Row: Compressed Convolution Networks for Graphs (ICML 2023 · similarity 0.92)
- Deeper Insights Into Graph Convolutional Networks for Semi-Supervised Learning (AAAI 2018 · similarity 0.91)
- L2-GCN: Layer-Wise and Learned Efficient Training of Graph Convolutional Networks (CVPR 2020 · similarity 0.91)
- Generalizing CNNs to graphs with learnable neighborhood quantization (NeurIPS 2024 · similarity 0.91)
- Towards Deepening Graph Neural Networks: A GNTK-based Optimization Perspective (ICLR 2022 · similarity 0.90)
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.