Incorporating spatial association into statistical classifiers: local pattern-based prior tuning

doi:10.1080/13658816.2020.1737702

	Incorporating spatial association into statistical classifiers: local pattern-based prior tuning
	Bai, Hexiang 3; Cao, Feng 3; Atkinson, M. Peter 1; Chen, Qian 3; Wang, Jinfeng 2; Ge, Yong 2
刊名	INTERNATIONAL JOURNAL OF GEOGRAPHICAL INFORMATION SCIENCE
	2020-03-12
页码	38
关键词	Spatial pattern statistical classifier spatial auto-logistic regression spatial data
ISSN号	1365-8816
DOI	10.1080/13658816.2020.1737702
通讯作者	Ge, Yong(gey@lreis.ac.cn)
英文摘要	This paper proposes a new classification method for spatial data by adjusting prior class probabilities according to local spatial patterns. First, the proposed method uses a classical statistical classifier to model training data. Second, the prior class probabilities are estimated according to the local spatial pattern and the classifier for each unseen object is adapted using the estimated prior probability. Finally, each unseen object is classified using its adapted classifier. Because the new method can be coupled with both generative and discriminant statistical classifiers, it performs generally more accurately than other methods for a variety of different spatial datasets. Experimental results show that this method has a lower prediction error than statistical classifiers that take no spatial information into account. Moreover, in the experiments, the new method also outperforms spatial auto-logistic regression and Markov random field-based methods when an appropriate estimate of local prior class distribution is used.
资助项目	Strategic Priority Research Program of the Chinese Academy of Science[XDA19040501] ; Chinese National Science Fundation[41725006] ; National Natural Science Foundation of China[41871286] ; National Natural Science Foundation of China[61672331] ; Natural Science Foundation of Shanxi Province, China[201701D121055]
WOS关键词	REMOTELY-SENSED IMAGES ; K-NN CLASSIFIER ; WEIGHTED REGRESSION ; COVER CHANGE ; EXTRACTION ; MODEL ; IDENTIFICATION ; INFORMATION ; PREDICTION ; EXPANSION
WOS研究方向	Computer Science ; Geography ; Physical Geography ; Information Science & Library Science
语种	英语
出版者	TAYLOR & FRANCIS LTD
WOS记录号	WOS:000519399200001
资助机构	Strategic Priority Research Program of the Chinese Academy of Science ; Chinese National Science Fundation ; National Natural Science Foundation of China ; Natural Science Foundation of Shanxi Province, China
内容类型	期刊论文
源URL	[http://ir.igsnrr.ac.cn/handle/311030/133016]
专题	中国科学院地理科学与资源研究所
通讯作者	Ge, Yong
作者单位	1.Univ Lancaster, Engn Bldg, Lancaster, England 2.Chinese Acad Sci, State Key Lab Resources & Environm Informat Syst, Inst Geog Sci & Nat Resources Res, Beijing, Peoples R China 3.Shanxi Univ, Sch Comp & Informat Technol, Taiyuan, Shanxi, Peoples R China
推荐引用方式 GB/T 7714	Bai, Hexiang,Cao, Feng,Atkinson, M. Peter,et al. Incorporating spatial association into statistical classifiers: local pattern-based prior tuning[J]. INTERNATIONAL JOURNAL OF GEOGRAPHICAL INFORMATION SCIENCE,2020:38.
APA	Bai, Hexiang,Cao, Feng,Atkinson, M. Peter,Chen, Qian,Wang, Jinfeng,&Ge, Yong.(2020).Incorporating spatial association into statistical classifiers: local pattern-based prior tuning.INTERNATIONAL JOURNAL OF GEOGRAPHICAL INFORMATION SCIENCE,38.
MLA	Bai, Hexiang,et al."Incorporating spatial association into statistical classifiers: local pattern-based prior tuning".INTERNATIONAL JOURNAL OF GEOGRAPHICAL INFORMATION SCIENCE (2020):38.