An Ensemble of Classifiers Based on Positive and Unlabeled Data in One-Class Remote Sensing Classification

doi:10.1109/JSTARS.2017.2789213

CORC > 植物研究所 > 中国科学院植物研究所 > 植被与环境变化国家重点实验室

	An Ensemble of Classifiers Based on Positive and Unlabeled Data in One-Class Remote Sensing Classification
	Liu, Ran ; Li, Wenkai ; Liu, Xiaoping ; Lu, Xingcheng 4; Li, Tianhong 1; Guo, Qinghua 3
刊名	IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING
	2018
卷号	11 期号:2 页码:572-584
关键词	Classifier ensemble one-class classification positive and unlabeled learning (PUL) weighted average weighted vote
ISSN号	1939-1404
DOI	10.1109/JSTARS.2017.2789213
文献子类	Article
英文摘要	One-class remote sensing classification refers to the situations when users are only interested in one specific land type without considering other types. The positive and unlabeled learning (PUL) algorithm, which trains a binary classifier from positive and unlabeled data, has been shown to be promising in one-class classification. The implementation of PUL by a single classifier has been investigated. However, implementing PUL using multiple classifiers and creating classifier ensembles based on PUL have not been studied. In this research, we investigate the implementations of PUL using several classifiers, including generalized linear model, generalized additive model, multivariate adaptive regression splines, maximum entropy, backpropagation neural network, and support vector machine, as well as three ensemble methods based on majority vote, weighted average, and weighted vote combination rules. These methods are applied in classifying the urban areas from four remote sensing imagery of different spatial resolutions, including aerial photograph, Landsat 8, WorldView-3, and Gaofen-1. Experimental results show that classifiers can successfully extract the urban areas with high accuracies, and the ensemble methods based on weighted average and weighted vote generally outperform the individual classifiers on different datasets. We conclude that PUL is a promising method in one-class remote sensing classification, and the classifier ensemble based on PUL can significantly improve the accuracy.
学科主题	Engineering, Electrical & Electronic ; Geography, Physical ; Remote Sensing ; Imaging Science & Photographic Technology
电子版国际标准刊号	2151-1535
出版地	PISCATAWAY
WOS关键词	MAXIMUM-ENTROPY APPROACH ; SUPPORT VECTOR MACHINES ; IMAGE CLASSIFICATION ; NEURAL-NETWORKS ; MODEL ; ACCURACY ; ALGORITHMS ; REGRESSION ; DIVERSITY
语种	英语
出版者	IEEE-INST ELECTRICAL ELECTRONICS ENGINEERS INC
WOS记录号	WOS:000425661700020
资助机构	National Natural Science Foundation of ChinaNational Natural Science Foundation of China (NSFC) [41401516] ; Fundamental Research Funds for the Central UniversitiesFundamental Research Funds for the Central Universities [15lgpy16]
内容类型	期刊论文
源URL	[http://ir.ibcas.ac.cn/handle/2S10CLM1/20492]
专题	植被与环境变化国家重点实验室
作者单位	1.Hong Kong Univ Sci & Technol, Div Environm, Hong Kong, Peoples R China 2.Chinese Acad Sci, Inst Bot, State Key Lab Vegetat & Environm Change, Beijing 100093, Peoples R China 3.Peking Univ, Coll Environm Sci & Engn, Beijing 100871, Peoples R China 4.Sun Yat Sen Univ, Sch Geog & Planning, Guangdong Prov Key Labo Urbanizat & Geosimulat, Guangzhou 510275, Guangdong, Peoples R China
推荐引用方式 GB/T 7714	Liu, Ran,Li, Wenkai,Liu, Xiaoping,et al. An Ensemble of Classifiers Based on Positive and Unlabeled Data in One-Class Remote Sensing Classification[J]. IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING,2018,11(2):572-584.
APA	Liu, Ran,Li, Wenkai,Liu, Xiaoping,Lu, Xingcheng,Li, Tianhong,&Guo, Qinghua.(2018).An Ensemble of Classifiers Based on Positive and Unlabeled Data in One-Class Remote Sensing Classification.IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING,11(2),572-584.
MLA	Liu, Ran,et al."An Ensemble of Classifiers Based on Positive and Unlabeled Data in One-Class Remote Sensing Classification".IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING 11.2(2018):572-584.