dc.creatorZimek, Arthur
dc.creatorCampello, Ricardo José Gabrielli Barreto
dc.creatorSander, Jörg
dc.date.accessioned2014-07-10T17:46:42Z
dc.date.accessioned2018-07-04T16:50:31Z
dc.date.available2014-07-10T17:46:42Z
dc.date.available2018-07-04T16:50:31Z
dc.date.created2014-07-10T17:46:42Z
dc.date.issued2014-06
dc.identifierInternational Conference on Scientific and Statistical Database Management, 26th, 2014, Aalborg.
dc.identifier9781450327220
dc.identifierhttp://www.producao.usp.br/handle/BDPI/45667
dc.identifierhttp://dx.doi.org/10.1145/2618243.2618257
dc.identifier.urihttp://repositorioslatinoamericanos.uchile.cl/handle/2250/1641103
dc.description.abstractOutlier detection and ensemble learning are well established research directions in data mining yet the application of ensemble techniques to outlier detection has been rarely studied. Building an ensemble requires learning of diverse models and combining these diverse models in an appropriate way. We propose data perturbation as a new technique to induce diversity in individual outlier detectors as well as a rank accumulation method for the combination of the individual outlier rankings in order to construct an outlier detection ensemble. In an extensive evaluation, we study the impact, potential, and shortcomings of this new approach for outlier detection ensembles. We show that this ensemble can significantly improve over weak performing base methods.
dc.languageeng
dc.publisherAssociation for Computing Machinery - ACM
dc.publisherAalborg University
dc.publisherAalborg
dc.relationInternational Conference on Scientific and Statistical Database Management, 26th
dc.rightsCopyright ACM
dc.rightsrestrictedAccess
dc.subjectoutlier detection
dc.subjectensemble
dc.titleData perturbation for outlier detection ensembles
dc.typeActas de congresos


Este ítem pertenece a la siguiente institución