Rank Aggregation for Pattern Classifier Selection in Remote Sensing Images

Página do item simplificado
dc.contributor.authorFaria, Fabio A.
dc.contributor.authorPedronette, Daniel C. G. [UNESP]
dc.contributor.authorSantos, Jefersson A. dos
dc.contributor.authorRocha, Anderson
dc.contributor.authorTorres, Ricardo da S.
dc.contributor.institutionUniversidade Estadual de Campinas (UNICAMP)
dc.contributor.institutionUniversidade Estadual Paulista (Unesp)
dc.contributor.institutionUniversidade Federal de Minas Gerais (UFMG)
dc.date.accessioned2014-12-03T13:11:26Z
dc.date.available2014-12-03T13:11:26Z
dc.date.issued2014-04-01
dc.description.abstractIn the past few years, segmentation and classification techniques have become a cornerstone of many successful remote sensing algorithms aiming at delineating geographic target objects. One common strategy relies on using multiple complex features to guide the delineation process with the objective of gathering complementary information for improving classification results. However, a persistent problem in this approach is how to combine different and noncorrelated feature descriptors automatically. In this regard, one solution is to combine them through multiple classifier systems (MCSs) in which the diversity of simple/non-complex classifiers is an essential issue in the definition of appropriate strategies for classifier fusion. In this paper, we propose a novel strategy for selecting classifiers (whereby a classifier is taken as a pair of learning method plus image descriptor) to be combined in MCS. In the proposed solution, diversity measures are used to assess the degree of agreement/disagreement between pairs of classifiers and ranked lists are created to sort them according to their diversity score. Thereafter, the classifiers are also sorted according to their performance through different evaluation measures (e. g., kappa and tau indices). In the end, a rank aggregation method is proposed to select the most suitable classifiers based on both the diversity and the effectiveness performance of classifiers. The proposed fusion framework has targeted at coffee crop classification and urban recognition but it is general enough to be used in a variety of other pattern recognition problems. Experimental results demonstrate that the novel strategy yields good results when compared to several baselines while using fewer classifiers and being much more efficient.en
dc.description.affiliationUniv Estadual Campinas, Inst Comp, BR-13083852 Sao Paulo, Brazil
dc.description.affiliationState Univ Sao Paulo UNESP, Dept Stat Appl Math & Comp, BR-13506900 Sao Paulo, Brazil
dc.description.affiliationUniv Fed Minas Gerais, Dept Comp Sci, BR-31270010 Belo Horizonte, MG, Brazil
dc.description.affiliationUnespState Univ Sao Paulo UNESP, Dept Stat Appl Math & Comp, BR-13506900 Sao Paulo, Brazil
dc.description.sponsorshipCoordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)
dc.description.sponsorshipConselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)
dc.description.sponsorshipFundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)
dc.description.sponsorshipMicrosoft Research
dc.description.sponsorshipIdCAPES: 1260-12-0
dc.description.sponsorshipIdCNPq: 306580/2012-8
dc.description.sponsorshipIdCNPq: 484254/2012-0
dc.description.sponsorshipIdCNPq: 304352/2012-8
dc.description.sponsorshipIdFAPESP: 10/14910-0
dc.description.sponsorshipIdFAPESP: 10/05647-4
dc.description.sponsorshipIdFAPESP: 12/18768-0
dc.description.sponsorshipIdFAPESP: 13/08645-0
dc.format.extent1103-1115
dc.identifierhttp://dx.doi.org/10.1109/JSTARS.2014.2303813
dc.identifier.citationIeee Journal Of Selected Topics In Applied Earth Observations And Remote Sensing. Piscataway: Ieee-inst Electrical Electronics Engineers Inc, v. 7, n. 4, p. 1103-1115, 2014.
dc.identifier.doi10.1109/JSTARS.2014.2303813
dc.identifier.issn1939-1404
dc.identifier.urihttp://hdl.handle.net/11449/113144
dc.identifier.wosWOS:000335390000010
dc.language.isoeng
dc.publisherInstitute of Electrical and Electronics Engineers (IEEE)
dc.relation.ispartofIeee Journal Of Selected Topics In Applied Earth Observations And Remote Sensing
dc.relation.ispartofjcr2.777
dc.relation.ispartofsjr1,547
dc.rights.accessRightsAcesso restrito
dc.sourceWeb of Science
dc.subjectCoffee crop classificationen
dc.subjectdiversity measuresen
dc.subjectinformation fusionen
dc.subjectmeta-learningen
dc.subjecturban recognitionen
dc.titleRank Aggregation for Pattern Classifier Selection in Remote Sensing Imagesen
dc.typeArtigo
dcterms.licensehttp://www.ieee.org/publications_standards/publications/rights/rights_policies.html
dcterms.rightsHolderIeee-inst Electrical Electronics Engineers Inc
unesp.campusUniversidade Estadual Paulista (Unesp), Instituto de Geociências e Ciências Exatas, Rio Claropt

Arquivos

Coleções

Artigos - Matemática - IGCE