Uso De Medidas De Complexidade Em Seleção De Atributos

dc.audience.educationlevelMestrado
dc.contributor.advisorLorena, Ana Carolina [UNIFESP]
dc.contributor.authorOkimoto, Lucas Chesini [UNIFESP]
dc.contributor.institutionUniversidade Federal de São Paulo (UNIFESP)pt
dc.date.accessioned2020-03-25T11:43:47Z
dc.date.available2020-03-25T11:43:47Z
dc.date.issued2018-07-31
dc.description.abstractFeature Selection Is An Important Pre-Processing Step Usually Mandatory In Data Analysis By Machine Learning Techniques. Its Objective Is To Reduce Data Dimensionality By Removing Irrelevant And Redundant Features From A Dataset. In This Work We Evaluate The Use Of Complexity Measures Of Classification Problems In Feature Selection (Fs). These Descriptors Allow Estimating The Intrinsic Difficulty Of A Classification Problem By Regarding On Characteristics Of The Dataset Available For Learning. We Propose A Combined Univariate-Multivariate Fs Technique Which Employs Two Of The Complexity Measures: Fisher "S Maximum Discriminant Ratio And Intra-Extra Class Distances. The Results Are Promising And Reveal That The Complexity Measures Are Indeed Suitable For Estimating Feature Importance In Classification Datasets. Large Reductions In The Numbers Of Features Were Obtained, While Preserving, In General, The Predictive Accuracy Of Two Strong Classification Techniques: Support Vector Machines And Random Forests.en
dc.description.abstractFeature Selection Is An Important Pre-Processing Step Usually Mandatory In Data Analysis By Machine Learning Techniques. Its Objective Is To Reduce Data Dimensionality By Removing Irrelevant And Redundant Features From A Dataset. In This Work We Evaluate The Use Of Complexity Measures Of Classification Problems In Feature Selection (Fs). These Descriptors Allow Estimating The Intrinsic Difficulty Of A Classification Problem By Regarding On Characteristics Of The Dataset Available For Learning. We Propose A Combined Univariate-Multivariate Fs Technique Which Employs Two Of The Complexity Measures: Fisher "S Maximum Discriminant Ratio And Intra-Extra Class Distances. The Results Are Promising And Reveal That The Complexity Measures Are Indeed Suitable For Estimating Feature Importance In Classification Datasets. Large Reductions In The Numbers Of Features Were Obtained, While Preserving, In General, The Predictive Accuracy Of Two Strong Classification Techniques: Support Vector Machines And Random Forests.pt
dc.description.sourceDados abertos - Sucupira - Teses e dissertações (2018)
dc.format.extent69 p.
dc.identifierhttps://sucupira.capes.gov.br/sucupira/public/consultas/coleta/trabalhoConclusao/viewTrabalhoConclusao.jsf?popup=true&id_trabalho=7085532pt
dc.identifier.file2018-0304.pdf
dc.identifier.urihttps://repositorio.unifesp.br/handle/11600/52365
dc.language.isoeng
dc.publisherUniversidade Federal de São Paulo (UNIFESP)
dc.rightsAcesso restrito
dc.subjectMachine Learningen
dc.subjectFeature Selectionen
dc.subjectDimensionalityen
dc.subjectComplexity Measuresen
dc.subjectClassificationen
dc.subjectClassificação Supervisionadapt
dc.subjectRedução De Dimensionalidadept
dc.subjectSeleção De Atributospt
dc.subjectMedidas De Complexidadept
dc.subjectAprendizado De Máquinapt
dc.titleUso De Medidas De Complexidade Em Seleção De Atributospt
dc.typeDissertação de mestrado
unifesp.campusSão José dos Campos, Instituto de Ciência e Tecnologiapt
unifesp.graduateProgramCiência Da Computaçãopt
unifesp.knowledgeAreaCiências Exatas E Da Terrapt
unifesp.researchAreaSistemas Inteligentespt
Arquivos
Pacote Original
Agora exibindo 1 - 1 de 1
Carregando...
Imagem de Miniatura
Nome:
dissertacao-lucas-okimoto_pos_revisao.pdf
Tamanho:
1.79 MB
Formato:
Adobe Portable Document Format
Descrição:
Dissertação Lucas Chesini Okimoto