Research map / Learning theory and algorithms / Clustering

Fast and Accurate $k$-means++ via Rejection Sampling

NeurIPS 2020 · Learning theory and algorithms › Clustering · approach: k-means++ seeding analysis

Read the paper Check 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 Fastk-means++ seeding instead of k-means++	k-means++ seeding · KDD-Cup, k=5000	running time ratio: 42.64x vs 1.0x	Helped
Use Fastk-means++ seeding instead of Afkmc2	k-means++ seeding · KDD-Cup, k=5000	running time ratio: 38.7x vs 1.0x	Helped
Use Fastk-means++ seeding instead of k-means++	k-means++ seeding · Song, k=5000	running time ratio: 46.26x vs 1.0x	Helped
Use Fastk-means++ seeding instead of Afkmc2	k-means++ seeding · Song, k=5000	running time ratio: 8.71x vs 1.0x	Helped
Use Fastk-means++ seeding instead of k-means++	k-means++ seeding · Census, k=5000	running time ratio: 36.69 vs 1.0x	Helped
Use Fastk-means++ seeding instead of Afkmc2	k-means++ seeding · Census, k=5000	running time ratio: 2.56x vs 1.0x	Helped
Use Fastk-means++ rather than uniform center sampling	k-means seeding · KDD-Cup, k=1000	k-means cost: 26199 -> 2957	Helped
Use Fastk-means++ rather than uniform center sampling	k-means seeding · Song, k=1000	k-means cost: 16373134 -> 14987614	Helped
Use Fastk-means++ rather than uniform center sampling	k-means seeding · Census, k=1000	k-means cost: 8678 -> 7883	Helped
Use Fastk-means++ rather than k-means++	k-means seeding · KDD-Cup, k=1000	k-means cost: 2695 -> 2957	Hurt
Use Fastk-means++ rather than k-means++	k-means seeding · Song, k=1000	k-means cost: 14746899 -> 14987614	Hurt
Use Fastk-means++ rather than k-means++	k-means seeding · Census, k=1000	k-means cost: 7812 -> 7883	Hurt

Closest papers by meaning

Approximate K-Means++ in Sublinear Time (AAAI 2016 · similarity 0.93)
A Better k-means++ Algorithm via Local Search (ICML 2019 · similarity 0.92)
k-variates++: more pluses in the k-means++ (ICML 2016 · similarity 0.91)
Multi-Swap k-Means++ (NeurIPS 2023 · similarity 0.89)
Near-Linear Time Approximation Algorithms for k-means with Outliers (ICML 2024 · similarity 0.89)
Analyzing $D^\alpha$ seeding for $k$-means (ICML 2024 · similarity 0.88)

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.

Open scime draft