期刊文献+

求解半定规划的ε-次微分向量丛方法

An ε-subgradient bundle algorithm for semidefinite programming
下载PDF
导出
摘要 本文基于ε 次微分向量丛理论和强对偶定理 ,通过寻求半定规划对偶问题的最优下降方向 ,得到原半定规划的最优值 .数值实验表明ε 次微分向量丛方法较适合于解大规模半定规划 . We get the optimal value of a semidefinite programming by f in ding the optimal descent direction of its dual problem based on the ε-subd ifferential bundle and strong duality. Numerical experiments indicate that this algorithm is effective for solving large-scale semidefinite programming.
出处 《应用数学》 CSCD 北大核心 2002年第1期108-112,共5页 Mathematica Applicata
基金 国家自然科学基金资助项目 (69972 0 36)
关键词 半定规划 ε-次微分向量丛 对偶问题 无约束非光滑优化问题 Semidefinite programming ε-subgradient bundle Duality
  • 相关文献

参考文献7

  • 1徐利治主编.现代数学手册(经济卷)[M].武汉:华中科技出版社,2001.
  • 2Oustry F. A second-order bundle method to minimize the maximum eigenvalue function[J]. Math. Program, 2000.89(1): 1~33.
  • 3Nesterov Y and Nemirovskii. A Interior point polynomial Algorithms in convex programming[A]. SIAM Studies in Applied Mathematics[C], Philadelphia, 1994.
  • 4Helmberg C and rendl F. A spectral bundle method for semidefinite programming[J]. SIAM J. Optim,1998, 10(3) :673~696.
  • 5Marko M, Makela and Pekka Neittaanmaki. Nonsmooth optimization: analysis and algorithms with application to optimization control[M]. Singapore: World scientific, 1992.
  • 6Helmberg C. Fixing variables in semidefinite relaxations[J]. SIAM J. Matrix Anal, 2000,21(3): 952~969.
  • 7Chih-Chien Tu, Ce-Kuen Shieh and Hsuanjen Cheng. Algorithms for graph partitioning problems by means of eigenspace relaxations[J]. European Journal of Operational Research,2000,123:86~104.

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