LEAST ABSOLUTE DEVIATIONS LEARNING OF MULTIPLE TASKS

doi:10.3934/jimo.2017071

CASIA OpenIR > 多模态人工智能系统全国重点实验室 > 人工智能与机器学习（杨雪冰）-技术团队

	LEAST ABSOLUTE DEVIATIONS LEARNING OF MULTIPLE TASKS
	Xue, Wei1,2 ; Zhang, Wensheng2,3 ; Yu, Gaohang 4
发表期刊	JOURNAL OF INDUSTRIAL AND MANAGEMENT OPTIMIZATION
ISSN	1547-5816
	2018-04-01
卷号	14 期号:2 页码:719-729
通讯作者	Xue, Wei(cswxue@ahut.edu.cn)
摘要	In this paper, we propose a new multitask feature selection model based on least absolute deviations. However, due to the inherent nonsmoothness of l(1) norm, optimizing this model is challenging. To tackle this problem efficiently, we introduce an alternating iterative optimization algorithm. Moreover, under some mild conditions, its global convergence result could be established. Experimental results and comparison with the state-of-the-art algorithm SLEP show the efficiency and effectiveness of the proposed approach in solving multitask learning problems.
关键词	Multitask learning feature selection least absolute deviations alternating direction method l(1) regularization
DOI	10.3934/jimo.2017071
关键词[WOS]	ALTERNATING DIRECTION METHOD ; SELECTION
收录类别	SCI
语种	英语
资助项目	NSFC[61305018] ; NSFC[61432008] ; NSFC[61472423] ; NSFC[61532006] ; NSFC[61262026] ; NSFC[11661007] ; Major Technologies R & D Special Program of Anhui, China[16030901060] ; NCET Programm of the Ministry of Education[NCET 13-0738] ; JGZX programm of Jiangxi Province[20112BCB23027] ; Natural Science Foundation of Jiangxi Province[20132BAB201-026] ; Science and Technology Programm of Jiangxi Education Committee[LDJH12-088] ; NSFC[61305018] ; NSFC[61432008] ; NSFC[61472423] ; NSFC[61532006] ; NSFC[61262026] ; NSFC[11661007] ; Major Technologies R & D Special Program of Anhui, China[16030901060] ; NCET Programm of the Ministry of Education[NCET 13-0738] ; JGZX programm of Jiangxi Province[20112BCB23027] ; Natural Science Foundation of Jiangxi Province[20132BAB201-026] ; Science and Technology Programm of Jiangxi Education Committee[LDJH12-088]
项目资助者	NSFC ; Major Technologies R & D Special Program of Anhui, China ; NCET Programm of the Ministry of Education ; JGZX programm of Jiangxi Province ; Natural Science Foundation of Jiangxi Province ; Science and Technology Programm of Jiangxi Education Committee
WOS研究方向	Engineering ; Operations Research & Management Science ; Mathematics
WOS类目	Engineering, Multidisciplinary ; Operations Research & Management Science ; Mathematics, Interdisciplinary Applications
WOS记录号	WOS:000438846500018
出版者	AMER INST MATHEMATICAL SCIENCES-AIMS
引用统计
文献类型	期刊论文
条目标识符	http://ir.ia.ac.cn/handle/173211/26335
专题	多模态人工智能系统全国重点实验室_人工智能与机器学习（杨雪冰）-技术团队
通讯作者	Xue, Wei
作者单位	1.Anhui Univ Technol, Sch Comp Sci & Technol, Maanshan 243032, Peoples R China 2.Nanjing Univ Sci & Technol, Sch Comp Sci & Engn, Nanjing 210094, Jiangsu, Peoples R China 3.Chinese Acad Sci, Inst Automat, Beijing 100190, Peoples R China 4.Gannan Normal Univ, Sch Math & Comp Sci, Ganzhou 341000, Peoples R China
推荐引用方式 GB/T 7714	Xue, Wei,Zhang, Wensheng,Yu, Gaohang. LEAST ABSOLUTE DEVIATIONS LEARNING OF MULTIPLE TASKS[J]. JOURNAL OF INDUSTRIAL AND MANAGEMENT OPTIMIZATION,2018,14(2):719-729.
APA	Xue, Wei,Zhang, Wensheng,&Yu, Gaohang.(2018).LEAST ABSOLUTE DEVIATIONS LEARNING OF MULTIPLE TASKS.JOURNAL OF INDUSTRIAL AND MANAGEMENT OPTIMIZATION,14(2),719-729.
MLA	Xue, Wei,et al."LEAST ABSOLUTE DEVIATIONS LEARNING OF MULTIPLE TASKS".JOURNAL OF INDUSTRIAL AND MANAGEMENT OPTIMIZATION 14.2(2018):719-729.