Evaluación de algoritmos de clasificación utilizando validación cruzada

Translated title of the contribution: Evaluation of classification algorithms using cross validation

Leticia Marisol Laura Ochoa

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Cross validation allows to evaluate the accuracy of the classification algorithms with low error, but has the problem of computational cost versus large volumes of data. In this work, the serial and parallel implementation of leave-one-out and k-fold cross validation techniques is performed using the R software environment. A comparison between the precision and error results obtained with cross validation techniques is presented, as well as the execution time thereof, reducing in the parallel implementation.

Translated title of the contributionEvaluation of classification algorithms using cross validation
Original languageSpanish
Title of host publication17th LACCEI International Multi-Conference for Engineering, Education, and Technology
Subtitle of host publication"Industry, Innovation, and Infrastructure for Sustainable Cities and Communities", LACCEI 2019
PublisherLatin American and Caribbean Consortium of Engineering Institutions
ISBN (Electronic)9780999344361
DOIs
StatePublished - 2019
Event17th LACCEI International Multi-Conference for Engineering, Education, and Technology, LACCEI 2019 - Montego Bay, Jamaica
Duration: 24 Jul 201926 Jul 2019

Publication series

NameProceedings of the LACCEI international Multi-conference for Engineering, Education and Technology
Volume2019-July
ISSN (Electronic)2414-6390

Conference

Conference17th LACCEI International Multi-Conference for Engineering, Education, and Technology, LACCEI 2019
Country/TerritoryJamaica
CityMontego Bay
Period24/07/1926/07/19

Bibliographical note

Funding Information:
Se aplicó esta técnica de validación cruzada de K iteraciones para copm arar los resultados de precisión y error de las técnicas de clasificación , así coom tiempos de ejecución . Se utilizaron cinco grupos, por lo que un grupo del total de datos se utilizó para la tabla de pruebas y los cuatro grupos restantes para la tabla de entrenamiento.

Publisher Copyright:
© 2019 Latin American and Caribbean Consortium of Engineering Institutions. All rights reserved.

Fingerprint

Dive into the research topics of 'Evaluation of classification algorithms using cross validation'. Together they form a unique fingerprint.

Cite this