期刊
MATHEMATICAL PROGRAMMING COMPUTATION
卷 12, 期 4, 页码 569-602出版社
SPRINGER HEIDELBERG
DOI: 10.1007/s12532-020-00177-4
关键词
Structured rank minimization; MPGCC; Exact penalty; Convex relaxation
类别
资金
- National Natural Science Foundation of China [11701186, 11971177, 11571120]
- Hong Kong Research Grant Council [PolyU153014/18p]
This paper concerns with a noisy structured low-rank matrix recovery problem which can be modeled as a structured rank minimization problem. We reformulate this problem as a mathematical program with a generalized complementarity constraint (MPGCC), and show that its penalty version, yielded by moving the generalized complementarity constraint to the objective, has the same global optimal solution set as the MPGCC does whenever the penalty parameter is over a certain threshold. Then, by solving the exact penalty problem in an alternating way, we obtain a multi-stage convex relaxation approach. We provide theoretical guarantees for our approach under a mild restricted eigenvalue condition, by quantifying the reduction of the error and approximate rank bounds of the first stage convex relaxation in the subsequent stages and establishing the geometric convergence of the error sequence in a statistical sense. Numerical experiments are conducted for some structured low-rank matrix recovery examples to confirm our theoretical findings. Our code can be achieved from 10.5281/zenodo.3600639.
作者
我是这篇论文的作者
点击您的名字以认领此论文并将其添加到您的个人资料中。
推荐
暂无数据