Model selection in discrete clustering: the EM-MML algorithm

Silvestre, Cláudia; Cardoso, Margarida; Figueiredo, Mário

Publication

Model selection in discrete clustering: the EM-MML algorithm

2016-12Conference object

dc.contributor.author	Silvestre, Cláudia
dc.contributor.author	Cardoso, Margarida
dc.contributor.author	Figueiredo, Mário
dc.date.accessioned	2017-12-12T14:48:58Z
dc.date.available	2017-12-12T14:48:58Z
dc.date.issued	2016-12
dc.description.abstract	Finite mixture models are widely used for cluster analysis in several areas of application. They are commonly estimated through likelihood maximization (using diverse variants of the expectation-maximization algorithm) and the number of components (or clusters) is determined resorting to information criteria: the EM algorithm is run several times and then one of the pre-estimated candidate models is selected (e.g. using the BIC criterion). We propose a new clustering approach to deal with the clustering of categorical data (quite common in social sciences) and simultaneously identify the number of clusters - the EM-MML algorithm. This approach assumes that the data comes from a finite mixture of multinomials and uses a variant of EM to estimate the model parameters and a minimum message length (MML) criterion to estimate the number of clusters. EM-MML thus seamlessly integrates estimation and model selection in a single algorithm. The EM-MML is compared with traditional EM approaches, using alternative information criteria. Comparisons rely on synthetic datasets and also on a real dataset (data from the European Social Survey). The results obtained illustrate the parsimony of the EM-MML solutions as well as their clusters cohesion-separation and stability. A clear advantage of EM-MML is also the computation time.	en
dc.description.version	N/A	pt_PT
dc.identifier.citation	SILVESTRE, Cláudia; CARDOSO, Margarida; FIGUEIREDO, Mário - Model selection in discrete clustering: the EM-MML algorithm. In: International Conference of the ERCIM WG on Computational and Methodological Statistics, 9th, Sevilha, Espanha, (Universidad de Sevilla), 2016 (9-11 de dezembro)	pt_PT
dc.identifier.uri	http://hdl.handle.net/10400.21/7688
dc.language.iso	eng	pt_PT
dc.peerreviewed	yes	pt_PT
dc.publisher	CMStatistics	pt_PT
dc.relation.publisherversion	http://cmstatistics.org/CMStatistics2016/fullprogramme.php	en
dc.relation.publisherversion	http://cmstatistics.org/RegistrationsV2/CMStatistics2016/viewSubmission.php?in=1307&token=q02p26s7q3nq46oo2422167os6856p69	en
dc.rights.uri	http://creativecommons.org/licenses/by-nc-sa/4.0/	pt_PT
dc.subject	Finite mixture models	en
dc.subject	EM-MML algorithm	en
dc.subject	Number of clusters	en
dc.title	Model selection in discrete clustering: the EM-MML algorithm	en
dc.type	conference object
dspace.entity.type	Publication
oaire.citation.conferencePlace	Universidad de Sevilla, Sevilla, Espanha	pt_PT
oaire.citation.title	: International Conference of the ERCIM WG on Computational and Methodological Statistics, 9th	en
person.familyName	Silvestre
person.givenName	Cláudia
person.identifier.ciencia-id	DA12-EF3F-C7CD
person.identifier.orcid	0000-0002-8850-4304
rcaap.rights	restrictedAccess	pt_PT
rcaap.type	conferenceObject	pt_PT
relation.isAuthorOfPublication	08fbc1bf-3387-4137-8c03-c4664dd43375
relation.isAuthorOfPublication.latestForDiscovery	08fbc1bf-3387-4137-8c03-c4664dd43375