Research map / Learning theory and algorithms / Clustering
Exact Recovery of Mangled Clusters with Same-Cluster Queries
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 ellipsoid-based recovery with monochromatic tessellation instead of centroid-based spherical recovery | Active clustering with same-cluster queries · Synthetic ellipsoidal clusters (n=10^5, k=5; d=2,4,6,8) | clustering error: 20%--40% -> exactly recovered | Helped |
| Use AlgoFULL rather than brute-force labeling | Synthetic ellipsoidal clusters (n=10^5, k=5; d=2,4,6,8) · Synthetic ellipsoidal clusters | queries to achieve clustering error ≤ .05: less than 3% of the queries needed by a brute force labeling | Helped |
| Add greedy hull expansion after sampling (ablation) | AlgoFULL · Synthetic ellipsoidal clusters | points recovered per round | Helped |
| Apply SCQ-k-means to clusters that are not centroid-centered spherical clusters | SCQ-k-means · Synthetic ellipsoidal clusters (n=10^5, k=5; d=2,4,6,8) | clustering error: 20%--40% | Hurt |
Closest papers by meaning
- On Margin-Based Cluster Recovery with Oracle Queries (NeurIPS 2021 · similarity 0.92)
- Clustering with Same-Cluster Queries (NeurIPS 2016 · similarity 0.90)
- Same-Cluster Querying for Overlapping Clusters (NeurIPS 2019 · similarity 0.88)
- Clustering with Non-adaptive Subset Queries (NeurIPS 2024 · similarity 0.87)
- Competitively Consistent Clustering (ICML 2025 · similarity 0.86)
- Optimal Algorithms for Learning Partitions with Faulty Oracles (NeurIPS 2024 · similarity 0.85)
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.