A Novel Manifold Regularized Online Semi-supervised Learning Model

doi:10.1007/s12559-017-9489-x

CORC > 数学与系统科学研究院 > 中国科学院数学与系统科学研究院 > 应用数学研究所

	A Novel Manifold Regularized Online Semi-supervised Learning Model
	Ding, Shuguang 1,2; Xi, Xuanyang 3; Liu, Zhiyong 3,4,5; Qiao, Hong 3,4,5; Zhang, Bo1,2
刊名	COGNITIVE COMPUTATION
	2018-02-01
卷号	10 期号:1 页码:49-61
关键词	Human learning Manifold regularization Online semi-supervised learning Lagrange dual problem
ISSN号	1866-9956
DOI	10.1007/s12559-017-9489-x
英文摘要	In the process of human learning, training samples are often obtained successively. Therefore, many human learning tasks exhibit online and semi-supervision characteristics, that is, the observations arrive in sequence and the corresponding labels are presented very sporadically. In this paper, we propose a novel manifold regularized model in a reproducing kernel Hilbert space (RKHS) to solve the online semi-supervised learning ((OSL)-L-2) problems. The proposed algorithm, named Model-Based Online Manifold Regularization (MOMR), is derived by solving a constrained optimization problem. Different from the stochastic gradient algorithm used for solving the online version of the primal problem of Laplacian support vector machine (LapSVM), the proposed algorithm can obtain an exact solution iteratively by solving its Lagrange dual problem. Meanwhile, to improve the computational efficiency, a fast algorithm is presented by introducing an approximate technique to compute the derivative of the manifold term in the proposed model. Furthermore, several buffering strategies are introduced to improve the scalability of the proposed algorithms and theoretical results show the reliability of the proposed algorithms. Finally, the proposed algorithms are experimentally shown to have a comparable performance to the standard batch manifold regularization algorithm.
资助项目	NSFC[61375005] ; NSFC[U1613213] ; NSFC[61210009] ; NSFC[61627808] ; NSFC[61603389] ; NSFC[61602483] ; MOST[2015BAK35B00] ; MOST[2015BAK35B01] ; Guangdong Science and Technology Department[2016B090910001] ; BNSF[4174107]
WOS研究方向	Computer Science ; Neurosciences & Neurology
语种	英语
出版者	SPRINGER
WOS记录号	WOS:000426075500006
内容类型	期刊论文
源URL	[http://ir.amss.ac.cn/handle/2S8OKBNM/29717]
专题	应用数学研究所
通讯作者	Liu, Zhiyong
作者单位	1.Chinese Acad Sci, LSEC, Beijing 100190, Peoples R China 2.Chinese Acad Sci, Inst Appl Math, AMSS, Beijing 100190, Peoples R China 3.Chinese Acad Sci, Inst Automat, State Key Lab Management & Control Complex Syst, Beijing 100190, Peoples R China 4.Chinese Acad Sci, CEBSIT, Shanghai 200031, Peoples R China 5.Chinese Acad Sci, Cloud Comp Ctr, Dongguan 523808, Guangdong, Peoples R China
推荐引用方式 GB/T 7714	Ding, Shuguang,Xi, Xuanyang,Liu, Zhiyong,et al. A Novel Manifold Regularized Online Semi-supervised Learning Model[J]. COGNITIVE COMPUTATION,2018,10(1):49-61.
APA	Ding, Shuguang,Xi, Xuanyang,Liu, Zhiyong,Qiao, Hong,&Zhang, Bo.(2018).A Novel Manifold Regularized Online Semi-supervised Learning Model.COGNITIVE COMPUTATION,10(1),49-61.
MLA	Ding, Shuguang,et al."A Novel Manifold Regularized Online Semi-supervised Learning Model".COGNITIVE COMPUTATION 10.1(2018):49-61.