截断高阶奇异值分解的增量算法

IF 1.6 3区数学 Q3 COMPUTER SCIENCE, SOFTWARE ENGINEERING

BIT Numerical Mathematics Pub Date : 2024-01-08 DOI:10.1007/s10543-023-01004-7

Chao Zeng, Michael K. Ng, Tai-Xiang Jiang

{"title":"截断高阶奇异值分解的增量算法","authors":"Chao Zeng, Michael K. Ng, Tai-Xiang Jiang","doi":"10.1007/s10543-023-01004-7","DOIUrl":null,"url":null,"abstract":"<p>We develop and study incremental algorithms for truncated higher-order singular value decompositions. By combining the SVD updating and different truncated higher-order singular value decompositions, two incremental algorithms are proposed. Not only the factor matrices but also the core tensor are updated in an incremental style. The costs of these algorithms are compared and the approximation errors are analyzed. Numerical results demonstrate that the proposed incremental algorithms have advantages in online computation.</p>","PeriodicalId":55351,"journal":{"name":"BIT Numerical Mathematics","volume":"154 1","pages":""},"PeriodicalIF":1.6000,"publicationDate":"2024-01-08","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Incremental algorithms for truncated higher-order singular value decompositions\",\"authors\":\"Chao Zeng, Michael K. Ng, Tai-Xiang Jiang\",\"doi\":\"10.1007/s10543-023-01004-7\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<p>We develop and study incremental algorithms for truncated higher-order singular value decompositions. By combining the SVD updating and different truncated higher-order singular value decompositions, two incremental algorithms are proposed. Not only the factor matrices but also the core tensor are updated in an incremental style. The costs of these algorithms are compared and the approximation errors are analyzed. Numerical results demonstrate that the proposed incremental algorithms have advantages in online computation.</p>\",\"PeriodicalId\":55351,\"journal\":{\"name\":\"BIT Numerical Mathematics\",\"volume\":\"154 1\",\"pages\":\"\"},\"PeriodicalIF\":1.6000,\"publicationDate\":\"2024-01-08\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"BIT Numerical Mathematics\",\"FirstCategoryId\":\"100\",\"ListUrlMain\":\"https://doi.org/10.1007/s10543-023-01004-7\",\"RegionNum\":3,\"RegionCategory\":\"数学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q3\",\"JCRName\":\"COMPUTER SCIENCE, SOFTWARE ENGINEERING\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"BIT Numerical Mathematics","FirstCategoryId":"100","ListUrlMain":"https://doi.org/10.1007/s10543-023-01004-7","RegionNum":3,"RegionCategory":"数学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"COMPUTER SCIENCE, SOFTWARE ENGINEERING","Score":null,"Total":0}

引用次数: 0

摘要

我们开发并研究了截断高阶奇异值分解的增量算法。通过结合 SVD 更新和不同的截断高阶奇异值分解，我们提出了两种增量算法。不仅因子矩阵，核心张量也以增量方式更新。对这些算法的成本进行了比较，并分析了近似误差。数值结果表明，所提出的增量算法在在线计算中具有优势。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

Incremental algorithms for truncated higher-order singular value decompositions

查看原文本刊更多论文

Incremental algorithms for truncated higher-order singular value decompositions

We develop and study incremental algorithms for truncated higher-order singular value decompositions. By combining the SVD updating and different truncated higher-order singular value decompositions, two incremental algorithms are proposed. Not only the factor matrices but also the core tensor are updated in an incremental style. The costs of these algorithms are compared and the approximation errors are analyzed. Numerical results demonstrate that the proposed incremental algorithms have advantages in online computation.

求助全文

通过发布文献求助，成功后即可免费获取论文全文。去求助

来源期刊

BIT Numerical Mathematics 数学-计算机：软件工程

CiteScore

2.90

自引率

0.00%

发文量

审稿时长

6 months

期刊介绍： The journal BIT has been published since 1961. BIT publishes original research papers in the rapidly developing field of numerical analysis. The essential areas covered by BIT are development and analysis of numerical methods as well as the design and use of algorithms for scientific computing. Topics emphasized by BIT include numerical methods in approximation, linear algebra, and ordinary and partial differential equations.