Research map / Graph learning / Graph algorithms and clustering
Certifying Robust Graph Classification under Orthogonal Gromov-Wasserstein Threats
Hongwei Jin, Zishun Yu, Xinhua Zhang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| resistance distance instead of shortest-path distance as base measure (ablation) | OGW SVM · MUTAG | accuracy: 76.1 -> 79.8 | Improved |
| resistance distance instead of shortest-path distance as base measure (ablation) | GW SVM · MUTAG | accuracy: 67.5 -> 72.1 | Improved |
| resistance distance instead of shortest-path distance as base measure (ablation) | OGW SVM · PTC-MR | accuracy: 57.9 -> 59.0 | Improved |
| resistance distance instead of shortest-path distance as base measure (ablation) | GW SVM · PTC-MR | accuracy: 56.1 -> 56.4 | Improved |
| resistance distance instead of shortest-path distance as base measure (ablation) | GW SVM · BZR | accuracy: 78.7 -> 79.0 | No clear effect |
| resistance distance instead of shortest-path distance as base measure (ablation) | OGW SVM · BZR | accuracy: 78.7 -> 78.7 | No clear effect |
| resistance distance instead of shortest-path distance as base measure (ablation) | GW and OGW SVM · COX2 | accuracy: 78.2 -> 78.2 | No clear effect |
| OGW instead of GW discrepancy (RD base) | SVM kernel · MUTAG | accuracy: 72.1 -> 79.8 | Improved |
| OGW instead of GW discrepancy (RD base) | SVM kernel · PTC-MR | accuracy: 56.4 -> 59.0 | Improved |
| OGW instead of GW discrepancy (RD base) | SVM kernel · BZR | accuracy: 79.0 -> 78.7 | No clear effect |
| OGW + RD SVM versus WL graph kernel | SVM kernel · PTC-MR | accuracy: 61.3 -> 59.0 | Hurt |
| resistance distance versus SP base measure in barycenters (ablation) | GW and OGW barycenter · synthetic cycle-like graphs | barycenter structure quality | No clear effect |
Closest papers by meaning
- Certified Robustness of Graph Convolution Networks for Graph Classification under Topological Attacks (NeurIPS 2020 · similarity 0.92)
- Robust Graph Condensation via Classification Complexity Mitigation (NeurIPS 2025 · similarity 0.89)
- Bounding the Expected Robustness of Graph Neural Networks Subject to Node Feature Attacks (ICLR 2024 · similarity 0.88)
- Deterministic Certification of Graph Neural Networks against Graph Poisoning Attacks with Arbitrary Perturbations (CVPR 2025 · similarity 0.87)
- On provable privacy vulnerabilities of graph representations (NeurIPS 2024 · similarity 0.86)
- Enhancing Graph Classification Robustness with Singular Pooling (NeurIPS 2025 · similarity 0.86)
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.
