浙江工业大学物理学院
 所在位置:首页 > 博学堂讲座
博学堂讲座
Gradient projection methods for the n-coupling problem (第275讲)
浏览量:1116    发布时间:2017-09-30 09:16:07

报告题目:Gradient projection methods for the n-coupling problem

报告人:Sangho Kum 教授

报告时间:上午10:30-11:30

报告地点:博C201

题目:Gradient projection methods for the n-coupling problem

报告人:Sangho Kum 教授(韩国忠北国立大学数学系(Chungbuk National University))

地点:浙工大屏峰校区博C201

时间:2017年10月03日  上午10:30-11:30

 

摘要:  We are concerned with optimization methods for the L2-Wasserstein
least squares problem of Gaussian measures (alternatively the n-coupling problem).
Based on its equivalent form on the convex cone of positive definite matrices
of fixed size and the strict convexity of the variance function, we are able
to present a (first) successful optimization algorithm for the unique minimizer.
Its global convergence rate analysis is provided according to the derived upper
bound of Lipschitz constants of the gradient function. 

报告人简介

Sangho Kum教授来自韩国忠北国立大学(Chungbuk National University).主要研究领域是: Convex Analysis and Optimization. 在国际重要学术期刊(SCI索引)Math. ProgramSIAM J. Optim上发表70余篇高质量的学术论文。(  Sangho Kum is a Professor of Department of Mathematics, Chungbuk National University of Korea. He received his PhD in Department of Mathematics  of  Seoul National University in 1991.  His research interests are mainly on convex analysis and optimization. )

 

博学堂讲座
Gradient projection methods for the n-coupling problem (第275讲)
浏览量:1116    发布时间:2017-09-30 09:16:07

报告题目:Gradient projection methods for the n-coupling problem

报告人:Sangho Kum 教授

报告时间:上午10:30-11:30

报告地点:博C201

题目:Gradient projection methods for the n-coupling problem

报告人:Sangho Kum 教授(韩国忠北国立大学数学系(Chungbuk National University))

地点:浙工大屏峰校区博C201

时间:2017年10月03日  上午10:30-11:30

 

摘要:  We are concerned with optimization methods for the L2-Wasserstein
least squares problem of Gaussian measures (alternatively the n-coupling problem).
Based on its equivalent form on the convex cone of positive definite matrices
of fixed size and the strict convexity of the variance function, we are able
to present a (first) successful optimization algorithm for the unique minimizer.
Its global convergence rate analysis is provided according to the derived upper
bound of Lipschitz constants of the gradient function. 

报告人简介

Sangho Kum教授来自韩国忠北国立大学(Chungbuk National University).主要研究领域是: Convex Analysis and Optimization. 在国际重要学术期刊(SCI索引)Math. ProgramSIAM J. Optim上发表70余篇高质量的学术论文。(  Sangho Kum is a Professor of Department of Mathematics, Chungbuk National University of Korea. He received his PhD in Department of Mathematics  of  Seoul National University in 1991.  His research interests are mainly on convex analysis and optimization. )