Research map / Graph learning / Graph algorithms and clustering
Wasserstein Embedding for Graph Learning
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 WEGL graph embeddings with AutoML instead of the listed GNN baselines | ogbg-molhiv · ogbg-molhiv scaffold split | Test ROC-AUC (%): 78.8 $\pm$ 0.8 -> 79.1 $\pm$ 0.3 | Helped |
| Add a virtual node to WEGL graphs (ablation) | WEGL + Random Forest · ogbg-molhiv scaffold split | Test ROC-AUC (%): 75.5 $\pm$ 1.5 -> 76.5 $\pm$ 1.8 | Helped |
| Replace WEGL with global average pooling (ablation) | WEGL + Virtual Node + Random Forest · ogbg-molhiv scaffold split | Test ROC-AUC (%): 76.5 $\pm$ 1.8 -> 72.1 $\pm$ 1.7 | Hurt |
| Use WEGL + Random Forest | IMDB-BINARY · IMDB-BINARY | accuracy (%): 76.9$\pm$3.6 -> 75.4$\pm$5.0 | Mixed |
| Use WEGL + GBDT | IMDB-MULTI · IMDB-MULTI | accuracy (%): 52.8$\pm$4.6 -> 52.3$\pm$2.9 | Mixed |
| Use WEGL + GBDT | COLLAB · COLLAB | accuracy (%): 83.6$\pm$1.0 -> 80.6$\pm$2.0 | Hurt |
| Use WEGL + GBDT | REDDIT-BINARY · REDDIT-BINARY | accuracy (%): 93.5$\pm$2.1 -> 92.9$\pm$1.9 | Mixed |
| Use WEGL + SVM-RBF | REDDIT-MULTI-5K · REDDIT-MULTI-5K | accuracy (%): 57.5$\pm$1.5 -> 56.1$\pm$2.3 | Mixed |
| Use WEGL + Random Forest | PTC-MR · PTC-MR | accuracy (%): 67.9$\pm$6.9 -> 67.5$\pm$7.7 | No clear effect |
| Use WEGL + Random Forest | ENZYMES · ENZYMES | accuracy (%): 59.1$\pm$1.1 -> 60.5$\pm$5.9 | Helped |
| Use WEGL + Random Forest | PROTEINS · PROTEINS | accuracy (%): 77.4$\pm$4.9 -> 76.5$\pm$4.2 | Mixed |
| Use a normal reference distribution instead of a k-means reference (ablation) | WEGL · ogbg-molhiv | Test ROC-AUC | No clear effect |
Closest papers by meaning
- Wasserstein Weisfeiler-Lehman Graph Kernels (NeurIPS 2019 · similarity 0.90)
- Deep Wasserstein Graph Discriminant Learning for Graph Classification (AAAI 2021 · similarity 0.90)
- Gromov-Wasserstein Learning for Graph Matching and Node Embedding (ICML 2019 · similarity 0.90)
- Weighted Embeddings for Low-Dimensional Graph Representation (AAAI 2025 · similarity 0.88)
- WatE: A Wasserstein t-distributed Embedding Method for Information-enriched Graph Visualization (AAAI 2025 · similarity 0.87)
- Learning Representations for Hierarchies with Minimal Support (NeurIPS 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.