A decomposition method for large-scale box constrained optimization

被引:39
|
作者
Yu, Jing [1 ]
Li, Mingqiang [2 ]
Wang, Yongli [2 ]
He, Guoping [2 ]
机构
[1] Beihang Univ, Sch Econ & Management, Beijing 100191, Peoples R China
[2] Shandong Univ Sci & Technol, Coll Informat Sci & Engn, Qingdao 266510, Peoples R China
基金
中国国家自然科学基金;
关键词
Box constrained optimization; Decomposition method; Working set identification; Large-scale; KKT-violating index; TRUST REGION ALGORITHMS; POINT NEWTON METHODS; STRICT COMPLEMENTARITY; CONVERGENCE; MINIMIZATION; BOUNDS;
D O I
10.1016/j.amc.2013.12.169
中图分类号
O29 [应用数学];
学科分类号
070104 ;
摘要
A decomposition method for solving large-scale box constrained optimization is proposed. The algorithm is motivated by the successful use of the decomposition method presented by Joachims for training support vector machines. In particular, a new technique, based on the new definition "KKT-violating index", is introduced for working set identification. Finally, the numerical experiments and implementation details show that this method is practical for large-scale problems. (C) 2014 Elsevier Inc. All rights reserved.
引用
收藏
页码:9 / 15
页数:7
相关论文
共 50 条