An Efficient Orthonormalization-Free Approach for Sparse Dictionary Learning and Dual Principal Component Pursuit

doi:10.3390/s20113041

CSpace

	An Efficient Orthonormalization-Free Approach for Sparse Dictionary Learning and Dual Principal Component Pursuit
	Hu, Xiaoyin 1,2; Liu, Xin 1,2,3
	2020-06-01
发表期刊	SENSORS
卷号	20 期号:11 页码:25
摘要	Sparse dictionary learning (SDL) is a classic representation learning method and has been widely used in data analysis. Recently, the lm-norm (m >= 3,m is an element of N) maximization has been proposed to solve SDL, which reshapes the problem to an optimization problem with orthogonality constraints. In this paper, we first propose an lm-norm maximization model for solving dual principal component pursuit (DPCP) based on the similarities between DPCP and SDL. Then, we propose a smooth unconstrained exact penalty model and show its equivalence with the lm-norm maximization model. Based on our penalty model, we develop an efficient first-order algorithm for solving our penalty model (PenNMF) and show its global convergence. Extensive experiments illustrate the high efficiency of PenNMF when compared with the other state-of-the-art algorithms on solving the lm-norm maximization with orthogonality constraints.
关键词	dual principal component pursuit orthogonality constraint sparse dictionary learning stiefel manifold
DOI	10.3390/s20113041
收录类别	SCI
语种	英语
资助项目	National Natural Science Foundation of China[11971466] ; National Natural Science Foundation of China[11991021] ; National Natural Science Foundation of China[11991020] ; Key Research Program of Frontier Sciences, Chinese Academy of Sciences[ZDBS-LY-7022] ; National Center for Mathematics and Interdisciplinary Sciences, Chinese Academy of Sciences ; Youth Innovation Promotion Association, Chinese Academy of Sciences
WOS研究方向	Chemistry ; Engineering ; Instruments & Instrumentation
WOS类目	Chemistry, Analytical ; Engineering, Electrical & Electronic ; Instruments & Instrumentation
WOS记录号	WOS:000552737900034
出版者	MDPI
引用统计
文献类型	期刊论文
条目标识符	http://ir.amss.ac.cn/handle/2S8OKBNM/51901
专题	中国科学院数学与系统科学研究院
通讯作者	Hu, Xiaoyin
作者单位	1.Chinese Acad Sci, Acad Math & Syst Sci, Beijing 100190, Peoples R China 2.Univ Chinese Acad Sci, Sch Math Sci, Beijing 100049, Peoples R China 3.State Key Lab Sci & Engn Comp, Beijing 100190, Peoples R China
推荐引用方式 GB/T 7714	Hu, Xiaoyin,Liu, Xin. An Efficient Orthonormalization-Free Approach for Sparse Dictionary Learning and Dual Principal Component Pursuit[J]. SENSORS,2020,20(11):25.
APA	Hu, Xiaoyin,&Liu, Xin.(2020).An Efficient Orthonormalization-Free Approach for Sparse Dictionary Learning and Dual Principal Component Pursuit.SENSORS,20(11),25.
MLA	Hu, Xiaoyin,et al."An Efficient Orthonormalization-Free Approach for Sparse Dictionary Learning and Dual Principal Component Pursuit".SENSORS 20.11(2020):25.