Author:
Fang Changjie,He Liangsong
Abstract
Abstract
Proximal algorithms has the advantages of low iteration cost and fast convergence speed, and is a common method for dealing with sparse structures. In this paper, we present an accelerated proximal extra-gradient method for figuring out the representative selection problems. We proved the convergence of the algorithm we proposed. Moreover, we apply our method to solve representative selection problem. Numerical experiments illustrate the advantages of our method.
Subject
General Physics and Astronomy
Reference16 articles.
1. Representative selection with structured sparsity;Wang;Pattern Recognition,2017
2. See all by looking at a few: Sparse modeling for finding representative objects;Elhamifar,2012
3. Nonlinear dictionary learning with application to image classification;Hu;Pattern Recognition,2018