Redundancy Is Not Necessarily Detrimental in Classification Problems

In feature selection, redundancy is one of the major concerns since the removal of redundancy 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 fe...

Descripción completa

Guardado en:

Detalles Bibliográficos
Autores principales:	Sebastián Alberto Grillo, José Luis Vázquez Noguera, Julio César Mello Román, Miguel García-Torres, Jacques Facon, Diego P. Pinto-Roa, Luis Salgueiro Romero, Francisco Gómez-Vela, Laura Raquel Bareiro Paniagua, Deysi Natalia Leguizamon Correa
Formato:	article
Lenguaje:	EN
Publicado:	MDPI AG 2021
Materias:	feature selection feature construction classification Mathematics QA1-939
Acceso en línea:	https://doaj.org/article/59081a7873574c42abfaf89268578daa
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

id	oai:doaj.org-article:59081a7873574c42abfaf89268578daa
record_format	dspace
spelling	oai:doaj.org-article:59081a7873574c42abfaf89268578daa2021-11-25T18:17:01ZRedundancy Is Not Necessarily Detrimental in Classification Problems10.3390/math92228992227-7390https://doaj.org/article/59081a7873574c42abfaf89268578daa2021-11-01T00:00:00Zhttps://www.mdpi.com/2227-7390/9/22/2899https://doaj.org/toc/2227-7390In feature selection, redundancy is one of the major concerns since the removal of redundancy 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.Sebastián Alberto GrilloJosé Luis Vázquez NogueraJulio César Mello RománMiguel García-TorresJacques FaconDiego P. Pinto-RoaLuis Salgueiro RomeroFrancisco Gómez-VelaLaura Raquel Bareiro PaniaguaDeysi Natalia Leguizamon CorreaMDPI AGarticlefeature selectionfeature constructionclassificationMathematicsQA1-939ENMathematics, Vol 9, Iss 2899, p 2899 (2021)
institution	DOAJ
collection	DOAJ
language	EN
topic	feature selection feature construction classification Mathematics QA1-939
spellingShingle	feature selection feature construction classification Mathematics QA1-939 Sebastián Alberto Grillo José Luis Vázquez Noguera Julio César Mello Román Miguel García-Torres Jacques Facon Diego P. Pinto-Roa Luis Salgueiro Romero Francisco Gómez-Vela Laura Raquel Bareiro Paniagua Deysi Natalia Leguizamon Correa Redundancy Is Not Necessarily Detrimental in Classification Problems
description	In feature selection, redundancy is one of the major concerns since the removal of redundancy 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.
format	article
author	Sebastián Alberto Grillo José Luis Vázquez Noguera Julio César Mello Román Miguel García-Torres Jacques Facon Diego P. Pinto-Roa Luis Salgueiro Romero Francisco Gómez-Vela Laura Raquel Bareiro Paniagua Deysi Natalia Leguizamon Correa
author_facet	Sebastián Alberto Grillo José Luis Vázquez Noguera Julio César Mello Román Miguel García-Torres Jacques Facon Diego P. Pinto-Roa Luis Salgueiro Romero Francisco Gómez-Vela Laura Raquel Bareiro Paniagua Deysi Natalia Leguizamon Correa
author_sort	Sebastián Alberto Grillo
title	Redundancy Is Not Necessarily Detrimental in Classification Problems
title_short	Redundancy Is Not Necessarily Detrimental in Classification Problems
title_full	Redundancy Is Not Necessarily Detrimental in Classification Problems
title_fullStr	Redundancy Is Not Necessarily Detrimental in Classification Problems
title_full_unstemmed	Redundancy Is Not Necessarily Detrimental in Classification Problems
title_sort	redundancy is not necessarily detrimental in classification problems
publisher	MDPI AG
publishDate	2021
url	https://doaj.org/article/59081a7873574c42abfaf89268578daa
work_keys_str_mv	AT sebastianalbertogrillo redundancyisnotnecessarilydetrimentalinclassificationproblems AT joseluisvazqueznoguera redundancyisnotnecessarilydetrimentalinclassificationproblems AT juliocesarmelloroman redundancyisnotnecessarilydetrimentalinclassificationproblems AT miguelgarciatorres redundancyisnotnecessarilydetrimentalinclassificationproblems AT jacquesfacon redundancyisnotnecessarilydetrimentalinclassificationproblems AT diegoppintoroa redundancyisnotnecessarilydetrimentalinclassificationproblems AT luissalgueiroromero redundancyisnotnecessarilydetrimentalinclassificationproblems AT franciscogomezvela redundancyisnotnecessarilydetrimentalinclassificationproblems AT lauraraquelbareiropaniagua redundancyisnotnecessarilydetrimentalinclassificationproblems AT deysinatalialeguizamoncorrea redundancyisnotnecessarilydetrimentalinclassificationproblems
_version_	1718411382107406336

Redundancy Is Not Necessarily Detrimental in Classification Problems

Ejemplares similares