Research map / Graph learning / Graph self-supervised learning
Substructure Assembling Network for Graph 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 SAU instead of LSTM unit (ablation) | SAN · MUTAG | accuracy: 80.21 -> 94.30 | Helped |
| use SAU instead of GRU unit (ablation) | SAN · MUTAG | accuracy: 81.33 -> 94.30 | Helped |
| use SAU instead of LSTM unit (ablation) | SAN · NCI1 | accuracy: 76.43 -> 83.84 | Helped |
| use SAU instead of GRU unit (ablation) | SAN · ENZYMES | accuracy: 53.33 -> 63.17 | Helped |
| use SAU instead of LSTM unit (ablation) | SAN · D&D | accuracy: 74.67 -> 81.37 | Helped |
| SAN vs Weisfeiler-Lehman kernel | SAN · NCI1 | accuracy: 84.55 -> 83.84 | No clear effect |
| SAN vs ECC | SAN · ENZYMES | accuracy: 53.50 -> 63.17 | Helped |
| SAN vs S2V | SAN · D&D | accuracy: 82.22 -> 81.37 | No clear effect |
| SAN vs PSCN | SAN · MUTAG | accuracy: 92.63 -> 94.30 | Helped |
| larger SAU dimensions (128,64,128) or more edge-selection layers (ablation) | SAN · NCI1 | accuracy | No clear effect |
| weighted average pooling instead of sum (ablation) | SAN · pooling | accuracy | No clear effect |
Closest papers by meaning
- Beyond Message Passing: Neural Graph Pattern Machine (ICML 2025 · similarity 0.88)
- Union Subgraph Neural Networks (AAAI 2024 · similarity 0.88)
- Implicit Subgraph Neural Network (ICML 2025 · similarity 0.88)
- A Flexible, Equivariant Framework for Subgraph GNNs via Graph Products and Graph Coarsening (NeurIPS 2024 · similarity 0.87)
- Deep Homomorphism Networks (NeurIPS 2024 · similarity 0.87)
- Translating Subgraphs to Nodes Makes Simple GNNs Strong and Efficient for Subgraph Representation Learning (ICML 2024 · 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.