Show simple item record  

dc.contributor.authorHunt, Lynette Anneen_NZ
dc.contributor.authorBasford, Kaye E.en_NZ
dc.date.accessioned2019-10-14T01:50:41Z
dc.date.available2016-11-01en_NZ
dc.date.available2019-10-14T01:50:41Z
dc.date.issued2016en_NZ
dc.identifier.citationHunt, L. A., & Basford, K. E. (2016). Comparing classical criteria for selecting intra-class correlated features in Multimix. Computational Statistics & Data Analysis, 103, 350–366. https://doi.org/10.1016/j.csda.2016.05.018en
dc.identifier.issn0167-9473en_NZ
dc.identifier.urihttps://hdl.handle.net/10289/12951
dc.description.abstractThe mixture approach to clustering requires the user to specify both the number of components to be fitted to the model and the form of the component distributions. In the Multimix class of models, the user also has to decide on the correlation structure to be introduced into the model. The behaviour of some commonly used model selection criteria is investigated when using the finite mixture model to cluster data containing mixed categorical and continuous attributes. The performance of these criteria in selecting both the number of components in the model and the form of the correlation structure amongst the attributes when fitting the Multimix class of models is illustrated using simulated data and a real medical data set. It is found that criteria based on the integrated classification likelihood have the best performance in detecting the number of clusters to be fitted to the model and in selecting the form of the component distributions. The performance of the Bayesian information criterion in detecting the correct model depends on the partitioning structure among the attributes while the Akaike information criterion and classification likelihood criterion perform in a less satisfactory way.
dc.format.mimetypeapplication/pdf
dc.language.isoenen_NZ
dc.publisherElsevieren_NZ
dc.rightsThis is an author’s accepted version of an article published in the journal: Computational Statistics & Data Analysis. © 2016 Elsevier.
dc.subjectScience & Technologyen_NZ
dc.subjectTechnologyen_NZ
dc.subjectPhysical Sciencesen_NZ
dc.subjectComputer Science, Interdisciplinary Applicationsen_NZ
dc.subjectStatistics & Probabilityen_NZ
dc.subjectComputer Scienceen_NZ
dc.subjectMathematicsen_NZ
dc.subjectModel selection criteriaen_NZ
dc.subjectFinite mixture modelsen_NZ
dc.subjectMixed dataen_NZ
dc.subjectMultimixen_NZ
dc.subjectDISCRIMINANT-ANALYSISen_NZ
dc.subjectCLUSTERING CRITERIAen_NZ
dc.subjectMIXTURE MODELen_NZ
dc.subjectNUMBERen_NZ
dc.subjectCOMPONENTSen_NZ
dc.subjectDISTANCEen_NZ
dc.subjectMachine learning
dc.titleComparing classical criteria for selecting intra-class correlated features in Multimixen_NZ
dc.typeJournal Article
dc.identifier.doi10.1016/j.csda.2016.05.018en_NZ
dc.relation.isPartOfComputational Statistics & Data Analysisen_NZ
pubs.begin-page350
pubs.elements-id139012
pubs.end-page366
pubs.publication-statusPublisheden_NZ
pubs.volume103en_NZ
dc.identifier.eissn1872-7352en_NZ


Files in this item

This item appears in the following Collection(s)

Show simple item record