{"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/smssvd-submatrix-selection-singular-value","title":"SMSSVD - SubMatrix Selection Singular Value Decomposition","arxiv_id":"1710.08144","date":"2017-10-23","proceeding":null,"authors":["Rasmus Henningsson","Magnus Fontes"],"abstract":"High throughput biomedical measurements normally capture multiple overlaid\nbiologically relevant signals and often also signals representing different\ntypes of technical artefacts like e.g. batch effects. Signal identification and\ndecomposition are accordingly main objectives in statistical biomedical\nmodeling and data analysis. Existing methods, aimed at signal reconstruction\nand deconvolution, in general, are either supervised, contain parameters that\nneed to be estimated or present other types of ad hoc features. We here\nintroduce SubMatrix Selection SingularValue Decomposition (SMSSVD), a\nparameter-free unsupervised signal decomposition and dimension reduction\nmethod, designed to reduce noise, adaptively for each low-rank-signal in a\ngiven data matrix, and represent the signals in the data in a way that enable\nunbiased exploratory analysis and reconstruction of multiple overlaid signals,\nincluding identifying groups of variables that drive different signals.\n  The Submatrix Selection Singular Value Decomposition (SMSSVD) method produces\na denoised signal decomposition from a given data matrix. The SMSSVD method\nguarantees orthogonality between signal components in a straightforward manner\nand it is designed to make automation possible. We illustrate SMSSVD by\napplying it to several real and synthetic datasets and compare its performance\nto golden standard methods like PCA (Principal Component Analysis) and SPC\n(Sparse Principal Components, using Lasso constraints). The SMSSVD is\ncomputationally efficient and despite being a parameter-free method, in\ngeneral, outperforms existing statistical learning methods.\n  A Julia implementation of SMSSVD is openly available on GitHub\n(https://github.com/rasmushenningsson/SMSSVD.jl).","url_abs":"http://arxiv.org/abs/1710.08144v1","url_pdf":"http://arxiv.org/pdf/1710.08144v1.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":"smssvd-submatrix-selection-singular-value","repo_url":"https://github.com/rasmushenningsson/SMSSVD.jl","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":null}],"tasks":[{"task_slug":"dimensionality-reduction","task_name":"Dimensionality Reduction"}],"methods":[{"method_slug":"pca","method_name":"PCA"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}