Research map / Learning theory and algorithms / Clustering
Robust and Fully-Dynamic Coreset for Continuous-and-Bounded Learning (With Outliers) Problems
Zixiu Wang, Yiwen Guo, Hu Ding
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| increase coreset size from 4e3 to 8e3 (ablation) | GSP+ · Covetype logistic regression | loss ratio: 1.046 -> 1.031 | Improved |
| increase coreset size from 4e3 to 8e3 (ablation) | GSP+ · Covetype logistic regression | speed-up: 26.9 -> 19.19 | Hurt |
| increase coreset size from 4e3 to 8e3 (ablation) | Uniform+ · Covetype logistic regression | loss ratio: 1.134 -> 1.050 | Improved |
| increase coreset size from 4e3 to 8e3 (ablation) | QR+ · Covetype logistic regression | loss ratio: 1.025 -> 1.012 | Improved |
| increase coreset size from 5e3 to 1e4 (ablation) | GSP+ · 3Dspatial k-means | loss ratio: 1.016 -> 1.008 | Improved |
| increase coreset size from 5e3 to 1e4 (ablation) | Uniform+ · 3Dspatial k-means | speed-up: 78.9 -> 46.9 | Hurt |
| use the QR+ black box instead of Uniform+ in the hybrid framework (ablation) | Robust coreset · Covetype logistic regression, |C|=4e3 | loss ratio: 1.134 -> 1.025 | Improved |
| add the hybrid outlier framework (+ suffix) to base coreset methods (ablation) | Uniform, GSP, QR coresets · Covetype and 3Dspatial | loss and accuracy | Improved |
| use coreset instead of rerunning on entire updated dataset | Fully-dynamic robust coreset · Dynamic setting | running time | Improved |
| increase tree height (smaller bucket size) in merge-and-reduce (ablation) | Dynamic robust coreset · Dynamic setting | speed-up | Improved |
Closest papers by meaning
- Robust Sparsification via Sensitivity (ICML 2025 · similarity 0.90)
- Coreset for Robust Geometric Median: Eliminating Size Dependency on Outliers (NeurIPS 2025 · similarity 0.89)
- The Impact of Coreset Selection on Spurious Correlations and Group Robustness (NeurIPS 2025 · similarity 0.86)
- Beyond Worst-Case Coreset Bounds for $k$-Clustering via Determinantal Sampling (NeurIPS 2026 · similarity 0.86)
- Universal Weak Coreset (AAAI 2024 · similarity 0.85)
- Coresets for Clustering Under Stochastic Noise (NeurIPS 2025 · 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.
