Exact penalization for cardinality and rank-constrained optimization problems via partial regularization

被引:0
|
作者
Lu, Zhaosong [1 ]
Li, Xiaorui [2 ]
Xiang, Shuhuang [3 ]
机构
[1] Univ Minnesota, Dept Ind & Syst Engn, Minneapolis, MN 55455 USA
[2] Huawei Technol Canada, Burnaby, BC, Canada
[3] Cent South Univ, Sch Math & Stat, Changsha, Hunan, Peoples R China
来源
OPTIMIZATION METHODS & SOFTWARE | 2023年 / 38卷 / 02期
基金
美国国家科学基金会;
关键词
Sparse optimization; low-rank optimization; cardinality constraint; rank constraint; partial regularization; exact penalty; VARIABLE SELECTION; CONVEX RELAXATION; SPARSE SIGNALS; MATRIX; MINIMIZATION; APPROXIMATION; OPTIMALITY;
D O I
10.1080/10556788.2022.2142583
中图分类号
TP31 [计算机软件];
学科分类号
081202 ; 0835 ;
摘要
In this paper, we consider a class of constrained optimization problems whose constraints involve a cardinality or rank constraint. The penalty formulation based on a partial regularization has recently been promoted in the literature to approximate these problems, which usually outperforms the penalty formulation based on a full regularization in terms of solution quality. Nevertheless, the relation between the penalty formulation with a partial regularizer and the original problem was not much studied yet. Under some suitable assumptions, we show that the penalty formulation based on a partial regularization is an exact reformulation of the original problem in the sense that they both share the same global minimizers. We also show that a local minimizer of the original problem is that of the penalty reformulation. These results provide some theoretical justification for the often-observed superior performance of the penalty model based on a partial regularizer over a corresponding full regularizer.
引用
收藏
页码:412 / 433
页数:22
相关论文
共 50 条