Mostrar el registro sencillo del ítem

dc.contributor.authorMillán Giraldo, Mónica
dc.contributor.authorSánchez Garreta, Josep Salvador
dc.contributor.authorTraver Roig, Vicente Javier
dc.date.accessioned2011-06-16T07:19:16Z
dc.date.available2011-06-16T07:19:16Z
dc.date.issued2009
dc.identifier.citationMILLÁN-GIRALDO, Mónica; SÁNCHEZ, J. Salvador; TRAVER, V. Javier. Exploring early classification strategies of streaming data with delayed attributes. En International Conference on Neural Information Processing. Springer, Berlin, Heidelberg, 2009. p. 875-883.
dc.identifier.citationMILLÁN-GIRALDO, Mónica; SÁNCHEZ, J. Salvador; TRAVER, V. Javier. Exploring early classification strategies of streaming data with delayed attributes. En International Conference on Neural Information Processing. Springer, Berlin, Heidelberg, 2009. p. 875-883.
dc.identifier.issn0302-9743
dc.identifier.urihttp://hdl.handle.net/10234/23621
dc.description.abstractIn contrast to traditional machine learning algorithms, where all data are available in batch mode, the new paradigm of streaming data poses additional difficulties, since data samples arrive in a sequence and many hard decisions have to be made on-line. The problem addressed here consists of classifying streaming data which not only are unlabeled, but also have a number l of attributes arriving after some time delay. In this context, the main issues are what to do when the unlabeled incomplete samples and, later on, their missing attributes arrive; when and how to classify these incoming samples; and when and how to update the training set. Three different strategies (for l = 1 and constant) are explored and evaluated in terms of the accumulated classification error. The results reveal that the proposed on-line strategies, despite their simplicity, may outperform classifiers using only the original, labeled-and-complete samples as a fixed training set. In other words, learning is possible by properly tapping into the unlabeled, incomplete samples, and their delayed attributes. The many research issues identified include a better understanding of the link between the inherent properties of the data set and the design of the most suitable on-line classification strategy
dc.format.extent10 p.
dc.language.isoeng
dc.publisherSpringer Verlag (Germany)
dc.relation.isFormatOfVersió pre-print del document publicat a: http://www.springerlink.com/content/61l730286324j336/
dc.relation.isPartOfLecture notes in computer science, 2009, v. 5863
dc.rights© Springer Verlag (Germany)
dc.rights.urihttp://www.springer.com/open+access/authors+rights?SGWID=0-176704-12-683201-0
dc.subjectData mining
dc.subjectStreaming data
dc.subjectOn-line classification
dc.subjectMissing attributes
dc.subject.lcshData mining
dc.subject.otherMineria de dades
dc.titleExploring early classification strategies of streaming data with delayed attributes
dc.typeinfo:eu-repo/semantics/article
dc.identifier.doihttp://dx.doi.org/10.1007/978-3-642-10677-4_99
dc.rights.accessRightsinfo:eu-repo/semantics/openAccess
dc.relation.publisherVersionhttps://link.springer.com/chapter/10.1007/978-3-642-10677-4_99
dc.relation.publisherVersionhttps://link.springer.com/chapter/10.1007/978-3-642-10677-4_99
dc.type.versioninfo:eu-repo/semantics/acceptedVersion


Ficheros en el ítem

Thumbnail

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

  • LSI_Articles [201]
    Articles de publicacions periòdiques escrits per professors del Departament de Llenguatges i Sistemes Informàtics

Mostrar el registro sencillo del ítem