期刊文献+
任意字段
题名或关键词
题名
关键词
文摘
作者
第一作者
机构
刊名
分类号
参考文献
作者简介
基金资助
栏目信息

求解极大相关问题的对偶方法

A Dual Method for the Maximal Correlation Problem
下载PDF
导出
摘要 多组变量间的极大相关问题(MCP)有重要统计应用。目前已有的求解MCP的算法都不能保证获得MCP的全局解。本文通过求解MCP的对偶问题,给出了一种改进的Lagrange对偶方法。最后,数值实验结果说明了新方法能提高收敛到全局解的可能性。 The maximal correlation problem aiming at assessing the relationship between sets of variables plays a very important role in many areas of statistical application.Currently,several algorithms for the global maximizer of the MCP have been proposed,but they can not guarantee convergence to a global solution.In the present paper,by solving the dual problem of the MCP,a modified Lagrange dual method is proposed.Numerical experiments demonstrate the new algorithm can increase the probability of finding aglobal maximizer.
作者 李美然 刘新国
机构地区 中国海洋大学数学科学学院
出处 《中国海洋大学学报（自然科学版）》 CAS CSCD 北大核心 2015年第8期128-132,共5页 Periodical of Ocean University of China
基金 国家自然科学基金项目(11371333)资助
关键词 极大相关问题 多元特征值问题 LAGRANGE对偶 强对偶 全局解 收敛性 maximal correlation problem multivariate eigenvalue problem lagrange dual strong dual global maximizer convergence
分类号 O212.4 [理学—概率论与数理统计]
  • 相关文献

参考文献17

  • 1Hotelling H. The most predictable criterion [J]. J Educ Pyschol,1935, 26: 139-142.
  • 2Hotelling H. Relations between two sets of variates [J]. Biometri- ka,1936, 28: 321-377.
  • 3Chu MT, Watterson J L. On a multivariate eigenvalue problem, Part I.. Algebraic theory and a power method [J]. SIAM J Sci Comput, 1993(14): 1089-1106.
  • 4HorstP. Relations among m sets of measures [J]. Psychometrika, 1961, 26: 129-149.
  • 5Hu D K. The convergence property of a algorithm about general- ized eigenvalue and eigenvector of positive definite matrix [C]. Bei- jing: China-Japan Symposium on Statistics, 1984.
  • 6孙继广.多参数特征值问题的一种算法[J].计算数学,1986,8(2):137-149.
  • 7秦晓伟,刘新国,赵娜.关于解极大相关问题P-SOR算法的收敛性[J].计算数学,2011,33(4):345-356. 被引量：3
  • 8Hanafi M, Ten Berge J M F. Global optimality of the successive Maxbet algorithm [J]. Psychometrika; 2003(68) : 97-103.
  • 9Zhang L H, Liao L Z. An alternating variable method for the max- imal correlation problem [J]. J Global Optim, DOI: 10. 1007/s 10898-011-9758-2.
  • 10Grzegorski S M. On the convergence of the method of alternating projections for multivariate symmetric eigenvalue problem [C]. Chania: Conference in Numerical Analysis, 2010: 15-18.

二级参考文献12

  • 1孙继广.多参数特征值问题的一种算法（I）[J].计算数学,1986,8(2):137-1490.
  • 2Hotelling H. The most predictable criterion[J]. J. Educ. Pyschol, 1935, 26, 139-142.
  • 3Hotelling H. Relations between two sets of variates[J]. Biometrika, 1936, 28, 321-377.
  • 4Horst P. Relations among m sets of measures[J]. Psychometrika, 1961, 26, 129-149.
  • 5Chu M T and Watterson J L. On a multivariate eigenvalue problem: I. Algebraic theory and power method[J]. SIAM J. Sci. Comput. 1993, 14, 1089-1106.
  • 6Hanafi M and Ten Berge J M F. Global optimality of the successive Maxbet algorithm[J]. Psychometrika, 2003, 68, 97-103.
  • 7Zhang L H and Chu M T. On a multivariate eigenvalue problem: II. Global solutions and the Gauss-Seidel method[J], available: http://www4.ncsu.edu/ mtchu.
  • 8Ostrowski A M. Solution of equations and systems of equations[M]. Academic press, New York, 1973.
  • 9Hu D K. The convergence property of a algorithm about generalized eigenvalue and eigenvector of positive definite matrix, China-Japan Symposium On Statistics, 1984, Beijing, China.
  • 10Bunch J R, Demmel J W and Van loan C F. The strong stability of algorithms for solving symmetric linear systems[J]. SIAM J. Matrix Anal. Appl. 1989, 10, 494-499.

共引文献2

中国海洋大学学报（自然科学版）
2015年 第8期

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部