Mostrar el registro sencillo del ítem

dc.contributor.authorGrillo, Sebastián Alberto 
dc.contributor.authorVázquez Noguera, José Luis 
dc.contributor.authorMello Román, Julio César
dc.contributor.authorGarcía Torres, Miguel
dc.contributor.authorFacon, Jacques
dc.contributor.authorPinto Roa, Diego Pedro 
dc.contributor.authorSalgueiro, Luis Fernando 
dc.contributor.authorBareiro Paniagua, Laura Raquel
dc.contributor.authorLeguizamón Correa, Deysi Natalia
dc.contributor.otherUniversidad Americana (PY)es
dc.contributor.otherUniversidad Nacional de Concepción (PY)es
dc.date.accessioned2022-04-29T23:02:18Z
dc.date.available2022-04-29T23:02:18Z
dc.date.issued2021
dc.identifier.urihttp://hdl.handle.net/20.500.14066/3777
dc.description.abstractIn feature selection, redundancy is one of the major concerns since the removal of redun dancy in data is connected with dimensionality reduction. Despite the evidence of such a connection, few works present theoretical studies regarding redundancy. In this work, we analyze the effect of redundant features on the performance of classification models. We can summarize the contribution of this work as follows: (i) develop a theoretical framework to analyze feature construction and selection, (ii) show that certain properly defined features are redundant but make the data linearly separable, and (iii) propose a formal criterion to validate feature construction methods. The results of experiments suggest that a large number of redundant features can reduce the classification error. The results imply that it is not enough to analyze features solely using criteria that measure the amount of information provided by such features.es
dc.description.sponsorshipCONACYT - Consejo Nacional de Ciencia y Tecnologíaes
dc.language.isoenges
dc.subject.classification1302 I+D en relación con la Ingenieríaes
dc.subject.otherFEATURE CONSTRUCTIONes
dc.subject.otherFEATURE SELECTIONes
dc.titleRedundancy Is Not Necessarily Detrimental in Classification Problemses
dc.typeresearch articlees
dc.description.fundingtextPROCIENCIAes
dc.journal.titleMathemathics Appliedes
dc.relation.projectCONACYTPINV18-1199es
dc.rights.accessRightsopen accesses


Ficheros en el ítem

Este ítem aparece en la(s) siguiente(s) colección(ones)

  • Artículos científicos
    La colección comprende artículos científicos, revisiones y artículos de conferencia que son resultados de actividades de I+D financiadas por el Programa PROCIENCIA.

Mostrar el registro sencillo del ítem