Research map / Graph learning / GNN architectures
Tree Mover's Distance: Bridging Graph Metrics and Stability of Graph Neural Networks
Ching-Yao Chuang, Stefanie Jegelka
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 |
|---|---|---|---|
| increase TMD depth L from 1 to 4 (ablation) | TMD-SVM · NCI1 | accuracy (%): 68.3 -> 84.8 | Improved |
| increase TMD depth L from 1 to 3 (ablation) | TMD-SVM · NCI109 | accuracy (%): 69.5 -> 82.3 | Improved |
| increase TMD depth L from 1 to 4 (ablation) | TMD-SVM · MUTAG | accuracy (%): 89.4 -> 92.2 | Improved |
| increase TMD depth L from 3 to 4 (ablation) | TMD-SVM · PTC | accuracy (%): 68.5 -> 66.5 | Hurt |
| increase TMD depth L from 3 to 4 (ablation) | TMD-SVM · COX2 | accuracy (%): 78.5 -> 76.1 | Hurt |
| increase TMD depth L from 1 to 3 (ablation) | TMD-SVM · BZR | accuracy (%): 83.8 -> 85.5 | Improved |
| increase TMD depth L from 1 to 4 (ablation) | TMD-SVM · PROTEINS | accuracy (%): 73.9 -> 75.2 | Improved |
| use TMD kernel SVM instead of WL subtree kernel | TMD L=4 · PTC | accuracy (%): 59.9 -> 66.5 | Improved |
| use TMD kernel SVM instead of Ramon & Gartner kernel | TMD L=4 · NCI1 | accuracy (%): 61.9 -> 84.8 | Improved |
| use TMD instead of WWL | TMD L=4 · NCI1 | accuracy (%): 86.1 -> 84.8 | Hurt |
| use TMD instead of GCN | TMD L=4 · NCI1 | accuracy (%): 80.2 -> 84.8 | Improved |
| parallelize tree OTs with 3 processes | TMD · DD | runtime (sec/pair): 32.07 -> 24.44 | Improved |
Closest papers by meaning
- Generalization, Expressivity, and Universality of Graph Neural Networks on Attributed Graphs (ICLR 2025 · similarity 0.91)
- Covered Forest: Fine-grained generalization analysis of graph neural networks (ICML 2025 · similarity 0.89)
- On the Rademacher Complexity of Graph Neural Networks: Unifying Expressivity and Geometry (NeurIPS 2026 · similarity 0.88)
- A Manifold Perspective on the Statistical Generalization of Graph Neural Networks (ICML 2025 · similarity 0.88)
- Generalization of Graph Neural Networks Is Robust to Model Mismatch (AAAI 2025 · similarity 0.88)
- Towards Bridging Generalization and Expressivity of Graph Neural Networks (ICLR 2025 · 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.
