Research map / Learning theory and algorithms / Clustering
Hierarchical Agglomerative Graph Clustering in Poly-Logarithmic Depth
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 eps=0.1 approximate instead of exact HAC (ablation) | ParHAC · UCI datasets | ARI: 1.5 | No clear effect |
| ParHAC_0.1 instead of SciPy average-linkage | pointset HAC · UCI datasets | ARI: 14.4 | Helped |
| ParHAC_0.1 instead of SciPy average-linkage | pointset HAC · UCI datasets | Dasgupta cost: 1.02 | Hurt |
| ParHAC_0.1 instead of best of SCC/Affinity | heuristic parallel HAC · UCI datasets | ARI: 35.6 | Helped |
| increase k in k-NN graph from 10 to 100 (ablation) | ParHAC eps=0.1 · digits | ARI: 10 | Hurt |
| increase k in k-NN graph from 10 to 1000 (ablation) | ParHAC eps=0.1 · digits | ARI: 47 | Hurt |
| larger epsilon (ablation) | ParHAC · large graphs | running time | Helped |
| different edge weighting schemes (6) (ablation) | ParHAC · large graphs | running time: 5x | No clear effect |
| ParHAC instead of SeqHAC | approximate HAC · large graphs | running time: 50.1x | Helped |
| ParHAC instead of Affinity | clustering · large graphs | running time: 14.8x | Hurt |
| ParHAC instead of fastcluster | pointset HAC · Glove-100 | end-to-end time: 417x | Helped |
Closest papers by meaning
- Hierarchical Agglomerative Graph Clustering in Nearly-Linear Time (ICML 2021 · similarity 0.95)
- Hierarchical Overlapping Clustering on Graphs: Cost Function, Algorithm and Scalability (ICML 2025 · similarity 0.88)
- New Bounds on the Cohesion of Complete-link and Other Linkage Methods for Agglomerative Clustering (ICML 2024 · similarity 0.88)
- Efficient Centroid-Linkage Clustering (NeurIPS 2024 · similarity 0.88)
- On the cohesion and separability of average-link for hierarchical agglomerative clustering (NeurIPS 2024 · similarity 0.87)
- From Logits to Hierarchies: Hierarchical Clustering made Simple (ICML 2025 · 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.