凸优化中的多面体逼近Polyhedral Approximations in Convex Optimization |
|
课程网址: | http://videolectures.net/opt08_bertsekas_paico/ |
主讲教师: | Dimitri Bertsekas |
开课单位: | 麻省理工学院 |
开课时间: | 2008-12-20 |
课程语种: | 英语 |
中文简介: | 我们提出了一种通过多面体近似求解凸程序的统一框架。它包括经典方法,如切割平面,Dantzig Wolfe分解,束和单纯形式组合,但也包括这些方法的改进,以及非常适合重要的大规模类型问题的新方法,例如在网络优化。 |
课程简介: | We propose a unifying framework for solution of convex programs by polyhedral approximation. It includes classical methods, such as cutting plane, Dantzig-Wolfe decomposition, bundle, and simplicial de- composition, but also includes refinements of these methods, as well as new methods that are well-suited for important large-scale types of problems, arising for example in network optimization. |
关 键 词: | 多面体; 凸程序; 切割平面 |
课程来源: | 视频讲座网 |
最后编审: | 2019-09-12:lxf |
阅读次数: | 94 |