Article ID | Journal | Published Year | Pages | File Type |
---|---|---|---|---|
4626287 | Applied Mathematics and Computation | 2015 | 8 Pages |
Abstract
In this paper, a fast proximal point algorithm (PPA) is proposed for solving ℓ1-minimization problem arising from compressed sensing. The proposed algorithm can be regarded as a new adaptive version of customized proximal point algorithm, which is based on a novel decomposition for the given nonsymmetric proximal matrix M. Since the proposed method is also a special case of the PPA-based contraction method, its global convergence can be established using the framework of a contraction method. Numerical results illustrate that the proposed algorithm outperforms some existing proximal point algorithms for sparse signal reconstruction.
Related Topics
Physical Sciences and Engineering
Mathematics
Applied Mathematics
Authors
Yun Zhu, Jian Wu, Gaohang Yu,