Research map / Learning theory and algorithms / Clustering
Multilevel Clustering via Wasserstein Means
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 MWMS instead of the strongest baseline, K-means | LabelMe image clustering · LabelMe | NMI: 0.349 -> 0.391 | Helped |
| Use MWMS instead of the strongest baseline, K-means | LabelMe image clustering · LabelMe | ARI: 0.237 -> 0.284 | Helped |
| Use MWMS instead of the strongest baseline, K-means | LabelMe image clustering · LabelMe | AMI: 0.324 -> 0.368 | Helped |
| Use MWMS instead of the fastest baseline, K-means | LabelMe image clustering · LabelMe | Time (s): 0.3 -> 544 | Hurt |
| Use MWM instead of the strongest baseline, K-means | LabelMe image clustering · LabelMe | NMI: 0.349 -> 0.373 | Helped |
| Use MWM instead of the strongest baseline, K-means | LabelMe image clustering · LabelMe | ARI: 0.237 -> 0.263 | Helped |
| Use MWM instead of the strongest baseline, K-means | LabelMe image clustering · LabelMe | AMI: 0.324 -> 0.352 | Helped |
| Use MWM instead of the fastest baseline, K-means | LabelMe image clustering · LabelMe | Time (s): 0.3 -> 332 | Hurt |
| Use Wasserstein multilevel clustering instead of three-stage K-means | Synthetic data with no sharing and constant variance · Synthetic data, no constraint and constant variance | Wasserstein distance to truth | Hurt |
| Use MWM or MWMS instead of three-stage K-means | Synthetic data with non-constant variance or shared local structure · Synthetic data, non-constant variance or local constraint | Wasserstein distance to truth | Helped |
Closest papers by meaning
- Wasserstein $K$-means for clustering probability distributions (NeurIPS 2022 · similarity 0.88)
- Relax and Merge: A Simple Yet Effective Framework for Solving Fair $k$-Means and $k$-sparse Wasserstein Barycenter Problems (ICLR 2025 · similarity 0.87)
- Expected Probabilistic Hierarchies (NeurIPS 2024 · similarity 0.86)
- Hierarchical Overlapping Clustering on Graphs: Cost Function, Algorithm and Scalability (ICML 2025 · similarity 0.86)
- Ultrametric Cluster Hierarchies: I Want ‘em All! (NeurIPS 2025 · similarity 0.86)
- THESAURUS: Contrastive Graph Clustering by Swapping Fused Gromov-Wasserstein Couplings (AAAI 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.