Research map / Graph learning / Graph self-supervised learning
Graph inference learning for semi-supervised classification
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 |
|---|---|---|---|
| Use GIL instead of DGCN | Semi-supervised node classification · Cora | accuracy: 83.5 -> 86.2 | Helped |
| Use GIL instead of DGCN | Semi-supervised node classification · Citeseer | accuracy: 72.6 -> 74.1 | Helped |
| Use GIL instead of DGCN | Semi-supervised node classification · Pubmed | accuracy: 80.0 -> 83.1 | Helped |
| Use GIL instead of DGCN | Semi-supervised node classification · NELL | accuracy: 74.2 -> 78.9 | Helped |
| Add meta-training from training nodes to validation nodes (ablation) | GIL trained on the training set · Cora | accuracy: 83.3 -> 86.2; gain of 2.9% | Helped |
| Jointly learn on training and validation sets (ablation) | GCN trained on the training set · Cora | accuracy: 81.4 -> 84.0; by 3.6% | Helped |
| Add structure-relation inference to node representations (ablation) | GIL with node representation only · Cora | accuracy: 81.5 -> 85.0 | Helped |
| Add path reachability (ablation) | GIL with node representation and structure relation · Cora | accuracy: 85.0 -> 86.2; 1.2% | Helped |
| Use node representation features (ablation) | GIL without modules · Cora | accuracy: 56.0 -> 81.5 | Helped |
| Use 2 convolutional layers instead of 1 (ablation) | GIL with mean pooling · Cora | accuracy: 84.5 -> 86.2; decreased by 1.7% with 1 layer | Helped |
| Use 2 convolutional layers instead of 3 (ablation) | GIL with mean pooling · Cora | accuracy: 85.4 -> 86.2 | Helped |
| Replace max pooling with mean pooling (ablation) | GIL with 2 convolutional layers · Cora | accuracy: 85.2 -> 86.2 | Helped |
Closest papers by meaning
- Semi-Supervised Classification with Graph Convolutional Networks (ICLR 2017 · similarity 0.91)
- A Flexible Generative Framework for Graph-based Semi-supervised Learning (NeurIPS 2019 · similarity 0.90)
- Hypergraph-enhanced Dual Semi-supervised Graph Classification (ICML 2024 · similarity 0.89)
- Unsupervised Neighborhood Propagation Kernel Layers for Semi-supervised Node Classification (AAAI 2024 · similarity 0.88)
- Node Identifiers: Compact, Discrete Representations for Efficient Graph Learning (ICLR 2025 · similarity 0.87)
- Fully-inductive Node Classification on Arbitrary Graphs (ICLR 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.