Research map / Graph learning / Graph algorithms and clustering
Optimal Transport for structured data with application on graphs
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 |
|---|---|---|---|
| combine node-feature distances and graph-structure distances in FGW | graph classification · vector-attributed graph datasets | accuracy | Helped |
| use Weisfeiler-Lehman node features instead of raw node labels (ablation) | FGW · MUTAG, NCI1, PTC-MR | accuracy: 83.26$\pm$10.30 -> 88.42$\pm$5.67; 72.82$\pm$1.46 -> 86.42$\pm$1.63; 55.71$\pm$6.74 -> 65.31$\pm$7.90 | Helped |
| increase WL neighborhood-label iterations from h=2 to h=4 (ablation) | FGW · MUTAG, NCI1, PTC-MR | accuracy: 86.42$\pm$7.81 -> 88.42$\pm$5.67; 85.82$\pm$1.16 -> 86.42$\pm$1.63; 63.20$\pm$7.68 -> 65.31$\pm$7.90 | Mixed |
| combine feature and structure information rather than using only features or only structure (ablation) | FGW · graph classification benchmarks | accuracy | Helped |
| use graph-structure comparison for unlabeled graphs | GW · IMDB-B and IMDB-M | accuracy: 56.00$\pm$3.61 -> 63.80$\pm$3.49; 41.13$\pm$4.68 -> 48.00$\pm$3.22 | Helped |
| use an intermediate feature-structure trade-off rather than alpha=0 or alpha=1 (ablation) | FGW · graph classification benchmarks | accuracy | Helped |
Closest papers by meaning
- Semi-relaxed Gromov-Wasserstein divergence and applications on graphs (ICLR 2022 · similarity 0.90)
- Bridging Graph Worlds: Neural Approximation of Gromov-Wasserstein Distances (NeurIPS 2026 · similarity 0.88)
- Template based Graph Neural Network with Optimal Transport Distances (NeurIPS 2022 · similarity 0.88)
- Wasserstein Weisfeiler-Lehman Graph Kernels (NeurIPS 2019 · similarity 0.88)
- Aligning Embeddings and Geometric Random Graphs: Informational Results and Computational Approaches for the Procrustes-Wasserstein Problem (NeurIPS 2024 · similarity 0.86)
- Generalized Sobolev Transport for Probability Measures on a Graph (ICML 2024 · 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.