Applied and Computational Harmonic Analysis ( IF 2.6 ) Pub Date : 2021-10-01 , DOI: 10.1016/j.acha.2021.09.003 Ning Bi 1, 2 , Wai-Shing Tang 3
In this paper, we focus on minimization model, i.e., investigating the nonconvex model: and provide a null space property of the measurement matrix A such that a vector x can be recovered from Ax via minimization. The minimization model was first proposed by E.Esser, et al (2013) [8]. As a nonconvex model, it is well known that global minimizer and local minimizer are usually inconsistent. In this paper, we present a necessary and sufficient condition for the measurement matrix A such that (1) a vector x can be recovered from Ax via local minimization (Theorem 4); (2) any k-sparse vector x can be recovered from Ax via local minimization (Theorem 5); (3) any k-sparse vector x can be recovered from Ax via global minimization (Theorem 6).
中文翻译:
通过ℓ1−ℓ2最小化进行稀疏向量恢复的充要条件
在本文中,我们专注于 最小化模型,即研究非凸模型:并提供测量矩阵的零空间属性甲使得向量X可从被回收斧经由最小化。这最小化模型首先由 E.Esser 等人 (2013) [8] 提出。作为非凸模型,众所周知,全局最小化器和局部最小化器通常是不一致的。在本文中,我们提出了一个充分必要条件测量矩阵甲使得(1)的向量X可从被回收斧经由局部最小化(定理 4);(2)任何ķ -sparse矢量X可以从回收的斧经由局部最小化(定理 5);(3)任何ķ -sparse矢量X可以从回收的斧经由 全局最小化(定理 6)。