An SVD-Based Algorithm for Orthogonal Low Rank Approximation of Tensors

发布者:王丹丹发布时间:2020-11-09浏览次数:416

学术报告(一)

  

报告题目:An SVD-Based Algorithm for Orthogonal Low Rank

          Approximation of Tensors

报告人:储德林教授 (新加坡国立大学)

报告时间:2020/11/12(周四)下午14:30-15:30

报告形式:腾讯会议

会议ID344 283 988

  

报告摘要:In this talk we consider orthogonal low rank approximation problem of tensors. We develop an SVD-based algorithm. Our SVD-based algorithm updates two vectors simultaneously and maintains the required orthogonality conditions by means of the polar decomposition. The convergence behaviour of our algorithm is analyzed for both objective function and iterates themselves. 

报告人简介:储德林,新加坡国立大学教授。1982年考入清华大学,获学士、硕士、博士学位。先后在香港大学,清华大学,德国TU Chemnitz(开姆尼斯工业大学)、University of Bielefeld(比勒费尔德大学)等高校工作过。任SIAM Journal on Scientific Computing, SIAM Journal on Matrix Analysis and Applications, Automatica期刊的副主编,Journal of Computational and Applied Mathematics的顾问编委。主要研究领域是科学计算、数据科学、数值代数及其应用,在SIAM系列杂志,Numerische MathematikMathematics of ComputationIEEE, Trans.Automatica 等国际知名学术期刊发表论文一百余篇。