Performance of first- and second-order methods for -regularized least squares problems

被引：0
|
作者：
Fountoulakis, Kimon [1 ,2 ]
Gondzio, Jacek [1 ,2 ,3 ]
机构：
[1] Univ Edinburgh, Sch Math, James Clerk Maxwell Bldg,Kings Bldg, Edinburgh EH9 3FD, Midlothian, Scotland
[2] Univ Edinburgh, Maxwell Inst, James Clerk Maxwell Bldg,Kings Bldg, Edinburgh EH9 3FD, Midlothian, Scotland
[3] NASK Res Inst, Wawozowa 18, PL-02796 Warsaw, Poland
来源：
COMPUTATIONAL OPTIMIZATION AND APPLICATIONS | 2016年 / 65卷 / 03期
关键词：
l(1)-Regularised least-squares; First-order methods; Second-order methods; Sparse least squares instance generator; Ill-conditioned problems; COORDINATE DESCENT METHOD; OPTIMIZATION; MINIMIZATION; CONVERGENCE; ALGORITHM;
D O I：
10.1007/s10589-016-9853-x
中图分类号：
C93 [管理学]; O22 [运筹学];
学科分类号：
070105 ; 12 ; 1201 ; 1202 ; 120202 ;
摘要：
We study the performance of first- and second-order optimization methods for -regularized sparse least-squares problems as the conditioning of the problem changes and the dimensions of the problem increase up to one trillion. A rigorously defined generator is presented which allows control of the dimensions, the conditioning and the sparsity of the problem. The generator has very low memory requirements and scales well with the dimensions of the problem.
引用
收藏
页码：605 / 635
页数：31
相关论文
共 50 条
12345