{"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/subsampled-online-matrix-factorization-with","title":"Subsampled online matrix factorization with convergence guarantees","arxiv_id":"1611.10041","date":"2016-11-30","proceeding":null,"authors":["Arthur Mensch","Julien Mairal","Gaël Varoquaux","Bertrand Thirion"],"abstract":"We present a matrix factorization algorithm that scales to input matrices\nthat are large in both dimensions (i.e., that contains morethan 1TB of data).\nThe algorithm streams the matrix columns while subsampling them, resulting in\nlow complexity per iteration andreasonable memory footprint. In contrast to\nprevious online matrix factorization methods, our approach relies on\nlow-dimensional statistics from past iterates to control the extra variance\nintroduced by subsampling. We present a convergence analysis that guarantees us\nto reach a stationary point of the problem. Large speed-ups can be obtained\ncompared to previous online algorithms that do not perform subsampling, thanks\nto the feature redundancy that often exists in high-dimensional settings.","url_abs":"http://arxiv.org/abs/1611.10041v1","url_pdf":"http://arxiv.org/pdf/1611.10041v1.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":"subsampled-online-matrix-factorization-with","repo_url":"https://github.com/arthurmensch/modl","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}