An online classification algorithm for large scale data streams: IGNGSVM

e-Archivo Repository

Show simple item record Suárez Cetrulo, Andrés L. Cervantes, Alejandro 2020-07-30T07:48:34Z 2020-07-30T07:48:34Z 2017-11-01
dc.identifier.bibliographicCitation Suárez-Cetrulo, A.L., Cervantes, A. (2017). An online classification algorithm for large scale data streams: iGNGSVM. Neurocomputing, 262, pp. 67-76.
dc.identifier.issn 0925-2312
dc.description.abstract Stream Processing has recently become one of the current commercial trends to face huge amounts of data. However, normally these techniques need specific infrastructures and high resources in terms of memory and computing nodes. This paper shows how mini-batch techniques and topology extraction methods can help making gigabytes of data to be manageable for just one server using computationally costly Machine Learning techniques as Support Vector Machines. The algorithm iGNGSVM is proposed to improve the performance of Support Vector Machines in datasets where the data is continuously arriving. It is benchmarked against a mini-batch version of LibSVM, achieving good accuracy rates and performing faster than this.
dc.language.iso eng
dc.publisher Elsevier
dc.rights © 2017 Elsevier B.V. All rights reserved.
dc.rights Atribución-NoComercial-SinDerivadas 3.0 España
dc.subject.other Data classification
dc.subject.other Topology extraction
dc.subject.other Online learning
dc.subject.other Large datasets
dc.subject.other Growing neural gas
dc.subject.other Support vector machines
dc.title An online classification algorithm for large scale data streams: IGNGSVM
dc.type article
dc.subject.eciencia Informática
dc.rights.accessRights openAccess
dc.type.version acceptedVersion
dc.identifier.publicationfirstpage 67
dc.identifier.publicationlastpage 76
dc.identifier.publicationtitle NEUROCOMPUTING
dc.identifier.publicationvolume 262
dc.identifier.uxxi AR/0000020150
 Find Full text

Files in this item

*Click on file's image for preview. (Embargoed files's preview is not supported)

The following license files are associated with this item:

This item appears in the following Collection(s)

Show simple item record