Research map / Learning theory and algorithms / Clustering
LSDS++ : Dual Sampling for Accelerated k-means++
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 |
|---|---|---|---|
| replace full LocalSearch++ swap evaluation with dual sampling (sample one random center plus nearest center) | LocalSearch++ step · DNA, RNA, PHY, BIO, MNIST (k=10,30) | k-means cost ratio vs k-means++ over iterations | No clear effect |
| dual sampling (LSDS++) versus LocalSearch++ with 500 steps then 20 local search | LSDS++ · five datasets, k=10 and k=30 | final cost ratio after 20 extra local search steps: within ~0.01 of LocalSearch++ in most cells; RNA k=10 lower (0.5541 vs 0.5633) | No clear effect |
| dual sampling per-iteration cost reduction (about factor k) relative to LocalSearch++ | LSDS++ · wall-clock time, k=10 and k=30 | running time to reach same cost: ~5x faster at k=10; ~10x faster at k=30 | Helped |
| LSDS++ followed by 20 local search steps vs k-means++ followed by 20 local search steps | LSDS++ · RNA k=10, PHY k=10, BIO k=10 | relative cost after 20 extra local search steps: RNA 0.5541 vs 0.6283; PHY 0.5950 vs 0.6674; BIO 0.5975 vs 0.6148 | Helped |
Closest papers by meaning
- Multi-Swap k-Means++ (NeurIPS 2023 · similarity 0.89)
- Fast and Accurate $k$-means++ via Rejection Sampling (NeurIPS 2020 · similarity 0.88)
- A Better k-means++ Algorithm via Local Search (ICML 2019 · similarity 0.88)
- A Constant-Factor Bi-Criteria Approximation Guarantee for k-means++ (NeurIPS 2016 · similarity 0.87)
- Near-Linear Time Approximation Algorithms for k-means with Outliers (ICML 2024 · similarity 0.87)
- Greedy Sampling for Approximate Clustering in the Presence of Outliers (NeurIPS 2019 · 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.