Classifier ensemble feature selection for automatic fault diagnosis

Página do item simplificado
dc.contributor.advisor-co1Varejão, Flávio Miguel
dc.contributor.advisor1Rauber, Thomas Walter
dc.contributor.authorBoldt, Francisco de Assis
dc.contributor.referee1Salles, Evandro Ottoni Teatini
dc.contributor.referee2Carvalho, André Carlos Ponce de Leon Ferreira de
dc.contributor.referee3Santos, Thiago Oliveira dos
dc.contributor.referee4Conci, Aura
dc.date.accessioned2018-08-02T00:04:07Z
dc.date.available2018-08-01
dc.date.available2018-08-02T00:04:07Z
dc.date.issued2017-07-14
dc.description.abstractAn efficient ensemble feature selection scheme applied for fault diagnosis is proposed, based on three hypothesis: a. A fault diagnosis system does not need to be restricted to a single feature extraction model, on the contrary, it should use as many feature models as possible, since the extracted features are potentially discriminative and the feature pooling is subsequently reduced with feature selection; b. The feature selection process can be accelerated, without loss of classification performance, combining feature selection methods, in a way that faster and weaker methods reduce the number of potentially non-discriminative features, sending to slower and stronger methods a filtered smaller feature set; c. The optimal feature set for a multi-class problem might be different for each pair of classes. Therefore, the feature selection should be done using an one versus one scheme, even when multi-class classifiers are used. However, since the number of classifiers grows exponentially to the number of the classes, expensive techniques like Error-Correcting Output Codes (ECOC) might have a prohibitive computational cost for large datasets. Thus, a fast one versus one approach must be used to alleviate such a computational demand. These three hypothesis are corroborated by experiments. The main hypothesis of this work is that using these three approaches together is possible to improve significantly the classification performance of a classifier to identify conditions in industrial processes. Experiments have shown such an improvement for the 1-NN classifier in industrial processes used as case study.eng
dc.description.resumoResumo
dc.formatText
dc.identifier.citationBOLDT, Francisco de Assis. Classifier ensemble feature selection for automatic fault diagnosis. 2017. 112 f. Tese (Doutorado em Ciência da Computação) - Universidade Federal do Espírito Santo, Centro Tecnológico, Vitória, 2017.
dc.identifier.urihttp://repositorio.ufes.br/handle/10/9872
dc.languageeng
dc.publisherUniversidade Federal do Espírito Santo
dc.publisher.countryBR
dc.publisher.courseDoutorado em Ciência da Computação
dc.publisher.departmentCentro Tecnológico
dc.publisher.initialsUFES
dc.publisher.programPrograma de Pós-Graduação em Informática
dc.rightsopen access
dc.subjectClassifier ensembleeng
dc.subjectFeature selectioneng
dc.subjectAutomatic fault diagnosiseng
dc.subjectSeleção de características (Computação)por
dc.subject.br-rjbnLocalização de falhas (Engenharia)
dc.subject.br-rjbnClassificadores (Linguistica)
dc.subject.cnpqCiência da Computação
dc.subject.udc004
dc.titleClassifier ensemble feature selection for automatic fault diagnosis
dc.typedoctoralThesis
Arquivos
Pacote Original
Agora exibindo 1 - 1 de 1
Imagem de Miniatura
Nome:
Doctor_thesis.pdf
Tamanho:
2.25 MB
Formato:
Adobe Portable Document Format
Descrição:
Baixar
Coleções
Doutorado em Ciência da Computação