Research map / Learning theory and algorithms / Clustering
Hierarchical Clustering Beyond the Worst-Case
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 SVD projection seeding (LINKAGE++) before single-linkage instead of plain linkage | LINKAGE++ · Synthetic HSBM | classification error: drops from 0.5 to 0 when n goes from 512 to 1024 (k=8) | Helped |
| Use SVD projection seeding followed by agglomerative merging (LINKAGE++) versus classic linkage heuristics | LINKAGE++ · Real-world sklearn datasets | Dasgupta cost (normalized) | Helped |
| Use only the SVD phase (PCA+) without the agglomerative phase on hierarchical data (ablation) | PCA+ · newsgroup (hierarchical) dataset | Dasgupta cost | Hurt |
| Use only phase 1 SVD projection (PCA+) on synthetic highly structured HSBM (ablation) | PCA+ · Synthetic HSBM | classification error and cost | No clear effect |
| Replace SVD seeding with single-linkage/complete-linkage heuristics alone | classic linkage heuristics · Synthetic HSBM | normalized Dasgupta cost | Hurt |
Closest papers by meaning
- Average Sensitivity of Hierarchical $k$-Median Clustering (ICML 2025 · similarity 0.91)
- Expected Probabilistic Hierarchies (NeurIPS 2024 · similarity 0.90)
- A Near-Linear Time Approximation Algorithm for Beyond-Worst-Case Graph Clustering (ICML 2024 · similarity 0.90)
- Ultrametric Cluster Hierarchies: I Want ‘em All! (NeurIPS 2025 · similarity 0.90)
- Hierarchical Overlapping Clustering on Graphs: Cost Function, Algorithm and Scalability (ICML 2025 · similarity 0.90)
- Learning-Augmented Hierarchical Clustering (ICML 2025 · similarity 0.89)
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.