{"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/the-k-means-u-algorithm-non-local-jumps-and","title":"The k-means-u* algorithm: non-local jumps and greedy retries improve k-means++ clustering","arxiv_id":"1706.09059","date":"2017-06-27","proceeding":null,"authors":["Bernd Fritzke"],"abstract":"We present a new clustering algorithm called k-means-u* which in many cases\nis able to significantly improve the clusterings found by k-means++, the\ncurrent de-facto standard for clustering in Euclidean spaces. First we\nintroduce the k-means-u algorithm which starts from a result of k-means++ and\nattempts to improve it with a sequence of non-local \"jumps\" alternated by runs\nof standard k-means. Each jump transfers the \"least useful\" center towards the\ncenter with the largest local error, offset by a small random vector. This is\ncontinued as long as the error decreases and often leads to an improved\nsolution. Occasionally k-means-u terminates despite obvious remaining\noptimization possibilities. By allowing a limited number of retries for the\nlast jump it is frequently possible to reach better local minima. The resulting\nalgorithm is called k-means-u* and dominates k-means++ wrt. solution quality\nwhich is demonstrated empirically using various data sets. By construction the\nlogarithmic quality bound established for k-means++ holds for k-means-u* as\nwell.","url_abs":"http://arxiv.org/abs/1706.09059v2","url_pdf":"http://arxiv.org/pdf/1706.09059v2.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":"the-k-means-u-algorithm-non-local-jumps-and","repo_url":"https://github.com/gittar/k-means-u-star","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"clustering","task_name":"Clustering"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}