Research map / Graph learning / GNN architectures
AdaGCN: Adaboosting Graph Convolutional Networks into Deep Models
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 |
|---|---|---|---|
| replace APPNP's fixed propagation weights with AdaGCN's adaptive AdaBoost combination of hop-specific predictions | APPNP-style graph prediction · Citeseer | accuracy: 75.41$\pm$0.35 -> 76.68$\pm$0.20 | Helped |
| replace APPNP's fixed propagation weights with AdaGCN's adaptive AdaBoost combination of hop-specific predictions | APPNP-style graph prediction · Cora-ML | accuracy: 84.28$\pm$0.28 -> 85.97$\pm$0.20 | Helped |
| replace APPNP's fixed propagation weights with AdaGCN's adaptive AdaBoost combination of hop-specific predictions | APPNP-style graph prediction · Pubmed | accuracy: 79.41$\pm$0.34 -> 79.95$\pm$0.21 | Helped |
| replace APPNP's fixed propagation weights with AdaGCN's adaptive AdaBoost combination of hop-specific predictions | APPNP-style graph prediction · MS Academic | accuracy: 92.98$\pm$0.07 -> 93.17$\pm$0.07; APPNP literature result 93.27$\pm$0.08 | Mixed |
| replace APPNP's fixed hop weights with adaptive AdaBoost weights | APPNP-style graph prediction · Reddit | F1-Score: 95.04$\pm$0.07 -> 95.39$\pm$0.13 | Helped |
| precompute hop-aggregated features and exclude sparse tensor operations from neural-network forward and backward passes | AdaGCN · Reddit | Per-epoch training time: 29489.81ms -> 32.29ms | Helped |
| increase the number of hop-specific base classifiers and combine their predictions (ablation) | AdaGCN · Citeseer, Cora-ML, Pubmed | accuracy | Helped |
| add residual connections to stacked GCN layers (ablation) | GCN · Citeseer, Cora-ML, Pubmed | accuracy | Mixed |
| remove ReLU and use simplified linear graph convolution (SGC) (ablation) | deep GCN · High-order neighbor prediction | accuracy | Mixed |
| replace AdaGCN's nonlinear base classifier with a linear transformation (AdaSGC) (ablation) | AdaGCN · High-order neighbor prediction | accuracy | Hurt |
| increase AdaGCN depth beyond the useful range on a high-label-rate, relatively simple graph (ablation) | AdaGCN · Reddit | F1-Score | Hurt |
Closest papers by meaning
- Beyond Graph Convolutional Network: An Interpretable Regularizer-Centered Optimization Framework (AAAI 2023 · similarity 0.90)
- Iterative Deep Graph Learning for Graph Neural Networks: Better and Robust Node Embeddings (NeurIPS 2020 · similarity 0.89)
- Generalizing CNNs to graphs with learnable neighborhood quantization (NeurIPS 2024 · similarity 0.89)
- Simplifying Graph Convolutional Networks (ICML 2019 · similarity 0.89)
- L2-GCN: Layer-Wise and Learned Efficient Training of Graph Convolutional Networks (CVPR 2020 · similarity 0.89)
- FastGCN: Fast Learning with Graph Convolutional Networks via Importance Sampling (ICLR 2018 · 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.