Improved analysis of SP and CoSaMP under total perturbations

Improved analysis of SP and CoSaMP under total perturbations
复制标题

改进了总扰动下 SP 和 CoSaMP 的分析

DOI:
10.1186/s13634-016-0412-5
复制
发表时间:
2016-11
影响因子:
1.9
通讯作者:
Li, Haifeng
Li, Haifeng
中科院分区:
工程技术4区
文献类型:
--
作者:
Li, Haifeng

文献摘要

参考文献

相似文献

实际上,在欠定模态中=Ax,其中,它不超过K个非零项),和A都可以被完全扰动。更宽松的条件意味着需要更少数量的测量来确保从理论方面的稀疏恢复。基于约束等距性(RIP),针对子空间追踪(SP)和压缩采样匹配追踪(CoSaMP),给出了在总扰动下保证稀疏向量轴恢复的两个放宽的充分条件.以随机矩阵作为测量矩阵,讨论了该条件的优越性。数值实验验证了SP和CoSaMP可以提供Oracle命令恢复性能。
Practically, in the underdetermined modely=Ax, wherexis aKsparse vector (i.e., it has no more thanKnonzero entries), bothyandAcould be totally perturbed. A more relaxed condition means less number of measurements are needed to ensure the sparse recovery from theoretical aspect. In this paper, based on restricted isometry property (RIP), for subspace pursuit (SP) and compressed sampling matching pursuit (CoSaMP), two relaxed sufficient conditions are presented under total perturbations to guarantee that the sparse vectorxis recovered. Taking random matrix as measurement matrix, we also discuss the advantage of our condition. Numerical experiments validate that SP and CoSaMP can provide oracle-order recovery performance.
DOI: --
发表时间: 2008
期刊: ArXiv
影响因子: --
作者:
Wei Dai;O. Milenkovic
通讯作者: Wei Dai;O. Milenkovic
DOI: 10.1016/j.acha.2008.07.002
发表时间: 2009-05-01
影响因子: 2.5
作者:
Needell, D.;Tropp, J. A.
通讯作者: Tropp, J. A.
DOI: 10.1109/tsp.2011.2174985
发表时间: 2012-03-01
影响因子: 5.4
作者:
Giryes, Raja;Elad, Michael
通讯作者: Elad, Michael
DOI: 10.1109/tit.2010.2048506
发表时间: 2010-07-01
影响因子: 2.5
作者:
Cai, Tony Tony;Wang, Lie;Xu, Guangwu
通讯作者: Xu, Guangwu
DOI: 10.1007/978-3-540-74494-8_43
发表时间: 2007-09
期刊: --
影响因子: --
作者:
T. Blumensath;M. Davies
通讯作者: T. Blumensath;M. Davies