{"title":"Skew-polynomial-sparse matrix multiplication","authors":"Qiao-Long Huang , Ke Ye , Xiao-Shan Gao","doi":"10.1016/j.jsc.2023.102240","DOIUrl":null,"url":null,"abstract":"<div><p>Based on the observation that <span><math><msup><mrow><mi>Q</mi></mrow><mrow><mo>(</mo><mi>p</mi><mo>−</mo><mn>1</mn><mo>)</mo><mo>×</mo><mo>(</mo><mi>p</mi><mo>−</mo><mn>1</mn><mo>)</mo></mrow></msup></math></span> is isomorphic to a quotient skew polynomial ring, we propose a new deterministic algorithm for <span><math><mo>(</mo><mi>p</mi><mo>−</mo><mn>1</mn><mo>)</mo><mo>×</mo><mo>(</mo><mi>p</mi><mo>−</mo><mn>1</mn><mo>)</mo></math></span> matrix multiplication over <span><math><mi>Q</mi></math></span>, where <em>p</em> is a prime number. The algorithm has complexity <span><math><mi>O</mi><mo>(</mo><msup><mrow><mi>T</mi></mrow><mrow><mi>ω</mi><mo>−</mo><mn>2</mn></mrow></msup><msup><mrow><mi>p</mi></mrow><mrow><mn>2</mn></mrow></msup><mo>)</mo></math></span>, where <span><math><mi>T</mi><mo>≤</mo><mi>p</mi><mo>−</mo><mn>1</mn></math></span> is a parameter determined by the skew-polynomial-sparsity of input matrices and <em>ω</em><span> is the asymptotic exponent of matrix multiplication<span>. Here a matrix is skew-polynomial-sparse if its corresponding skew polynomial is sparse. Moreover, by introducing randomness, we also propose a probabilistic algorithm with complexity </span></span><span><math><msup><mrow><mi>O</mi></mrow><mrow><mo>∼</mo></mrow></msup><mo>(</mo><msup><mrow><mi>t</mi></mrow><mrow><mi>ω</mi><mo>−</mo><mn>2</mn></mrow></msup><msup><mrow><mi>p</mi></mrow><mrow><mn>2</mn></mrow></msup><mo>+</mo><msup><mrow><mi>p</mi></mrow><mrow><mn>2</mn></mrow></msup><mi>log</mi><mo></mo><mfrac><mrow><mn>1</mn></mrow><mrow><mi>ν</mi></mrow></mfrac><mo>)</mo></math></span>, where <span><math><mi>t</mi><mo>≤</mo><mi>p</mi><mo>−</mo><mn>1</mn></math></span> is the skew-polynomial-sparsity of the product and <em>ν</em><span> is the probability parameter. The main feature of the algorithms is the acceleration for matrix multiplication if the input matrices or their products are skew-polynomial-sparse.</span></p></div>","PeriodicalId":0,"journal":{"name":"","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2023-07-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"","FirstCategoryId":"100","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0747717123000548","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
Based on the observation that is isomorphic to a quotient skew polynomial ring, we propose a new deterministic algorithm for matrix multiplication over , where p is a prime number. The algorithm has complexity , where is a parameter determined by the skew-polynomial-sparsity of input matrices and ω is the asymptotic exponent of matrix multiplication. Here a matrix is skew-polynomial-sparse if its corresponding skew polynomial is sparse. Moreover, by introducing randomness, we also propose a probabilistic algorithm with complexity , where is the skew-polynomial-sparsity of the product and ν is the probability parameter. The main feature of the algorithms is the acceleration for matrix multiplication if the input matrices or their products are skew-polynomial-sparse.