Research map / Learning theory and algorithms / Clustering
Near-optimal Coresets for Robust Clustering
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 the proposed coreset to run LL instead of on the full dataset | LL for robust Means · Adult, Bank, Twitter, Census1990 | running time: 80x-250x | Helped |
| Use the proposed coreset to run LS instead of on the full dataset | LS for robust Median · Adult, Bank, Twitter, Census1990 | running time: 100x-200x | Helped |
| Include the coreset construction time when running LL on the coreset | LL for robust Means · Adult, Bank, Twitter, Census1990 | running time: 10x-30x | Helped |
| Include the coreset construction time when running LS on the coreset | LS for robust Median · Adult, Bank, Twitter, Census1990 | running time: 80x-140x | Helped |
| Use the proposed coreset rather than uniform sampling | Robust Median coreset construction · All four datasets and tested target sizes | empirical error | Helped |
| Use the proposed coreset rather than sensitivity sampling | Robust Median coreset construction · All four datasets and tested target sizes | empirical error | Helped |
| Add the initial outliers to uniform sampling (OAUS instead of US) (ablation) | Uniform-sampling coreset baseline · All tested datasets | empirical error | No clear effect |
| Increase the number of outliers while keeping N - m fixed (ablation) | Proposed coreset · Robust Median experiments | empirical error | No clear effect |
| Use the proposed coreset to run LL instead of on the full dataset | LL for robust Means · Adult, Bank, Twitter, Census1990 | clustering cost: $3.790\times 10^{13}$ -> $3.922\times 10^{13}$; $4.444\times 10^8$ -> $4.652\times 10^8$; $3.218\times 10^7$ -> $3.236\t | Mixed |
| Use the proposed coreset to run LS instead of on the full dataset | LS for robust Median · Adult, Bank, Twitter, Census1990 | clustering cost: $1.100\times 10^9$ -> $1.107\times 10^9$; $4.717\times 10^6$ -> $4.721\times 10^6$; $1.476\times 10^6$ -> $1.451 \times | Mixed |
Closest papers by meaning
- Coreset for Robust Geometric Median: Eliminating Size Dependency on Outliers (NeurIPS 2025 · similarity 0.92)
- Beyond Worst-Case Coreset Bounds for $k$-Clustering via Determinantal Sampling (NeurIPS 2026 · similarity 0.92)
- Coresets for Clustering Under Stochastic Noise (NeurIPS 2025 · similarity 0.91)
- Universal Weak Coreset (AAAI 2024 · similarity 0.90)
- Improved Guarantees for Fully Dynamic $k$-Center Clustering with Outliers in General Metric Spaces (NeurIPS 2024 · similarity 0.90)
- Competitively Consistent 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.