{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/scalable-k-means-clustering-via-lightweight","title":"Scalable k-Means Clustering via Lightweight Coresets","arxiv_id":"1702.08248","date":"2017-02-27","proceeding":null,"authors":["Olivier Bachem","Mario Lucic","Andreas Krause"],"abstract":"Coresets are compact representations of data sets such that models trained on\na coreset are provably competitive with models trained on the full data set. As\nsuch, they have been successfully used to scale up clustering models to massive\ndata sets. While existing approaches generally only allow for multiplicative\napproximation errors, we propose a novel notion of lightweight coresets that\nallows for both multiplicative and additive errors. We provide a single\nalgorithm to construct lightweight coresets for k-means clustering as well as\nsoft and hard Bregman clustering. The algorithm is substantially faster than\nexisting constructions, embarrassingly parallel, and the resulting coresets are\nsmaller. We further show that the proposed approach naturally generalizes to\nstatistical k-means clustering and that, compared to existing results, it can\nbe used to compute smaller summaries for empirical risk minimization. In\nextensive experiments, we demonstrate that the proposed algorithm outperforms\nexisting data summarization strategies in practice.","url_abs":"http://arxiv.org/abs/1702.08248v2","url_pdf":"http://arxiv.org/pdf/1702.08248v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"scalable-k-means-clustering-via-lightweight","repo_url":"https://github.com/webis-de/small-text","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"clustering","task_name":"Clustering"},{"task_slug":"data-summarization","task_name":"Data Summarization"}],"methods":[{"method_slug":"k-means-clustering","method_name":"k-Means Clustering"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}