An alternative classification to mixture modeling for longitudinal counts or binary measures

Hdl Handle:
http://hdl.handle.net/10144/336371
Title:
An alternative classification to mixture modeling for longitudinal counts or binary measures
Authors:
Subtil, F; Boussari, O; Bastard, M; Etard, J-F; Ecochard, R; Génolini, C
Journal:
Statistical Methods in Medical Research
Abstract:
Classifying patients according to longitudinal measures, or trajectory classification, has become frequent in clinical research. The k-means algorithm is increasingly used for this task in case of continuous variables with standard deviations that do not depend on the mean. One feature of count and binary data modeled by Poisson or logistic regression is that the variance depends on the mean; hence, the within-group variability changes from one group to another depending on the mean trajectory level. Mixture modeling could be used here for classification though its main purpose is to model the data. The results obtained may change according to the main objective. This article presents an extension of the k-means algorithm that takes into account the features of count and binary data by using the deviance as distance metric. This approach is justified by its analogy with the classification likelihood. Two applications are presented with binary and count data to show the differences between the classifications obtained with the usual Euclidean distance versus the deviance distance.
Publisher:
SAGE Publications
Issue Date:
1-Sep-2014
URI:
http://hdl.handle.net/10144/336371
DOI:
10.1177/0962280214549040
PubMed ID:
25179548
Language:
en
ISSN:
1477-0334
Appears in Collections:
Research Methods

Full metadata record

DC FieldValue Language
dc.contributor.authorSubtil, Fen_GB
dc.contributor.authorBoussari, Oen_GB
dc.contributor.authorBastard, Men_GB
dc.contributor.authorEtard, J-Fen_GB
dc.contributor.authorEcochard, Ren_GB
dc.contributor.authorGénolini, Cen_GB
dc.date.accessioned2014-11-30T15:18:48Z-
dc.date.available2014-11-30T15:18:48Z-
dc.date.issued2014-09-01-
dc.identifier.citationAn alternative classification to mixture modeling for longitudinal counts or binary measures. 2014: Stat Methods Med Resen_GB
dc.identifier.issn1477-0334-
dc.identifier.pmid25179548-
dc.identifier.doi10.1177/0962280214549040-
dc.identifier.urihttp://hdl.handle.net/10144/336371-
dc.description.abstractClassifying patients according to longitudinal measures, or trajectory classification, has become frequent in clinical research. The k-means algorithm is increasingly used for this task in case of continuous variables with standard deviations that do not depend on the mean. One feature of count and binary data modeled by Poisson or logistic regression is that the variance depends on the mean; hence, the within-group variability changes from one group to another depending on the mean trajectory level. Mixture modeling could be used here for classification though its main purpose is to model the data. The results obtained may change according to the main objective. This article presents an extension of the k-means algorithm that takes into account the features of count and binary data by using the deviance as distance metric. This approach is justified by its analogy with the classification likelihood. Two applications are presented with binary and count data to show the differences between the classifications obtained with the usual Euclidean distance versus the deviance distance.en_GB
dc.languageENG-
dc.language.isoenen
dc.publisherSAGE Publicationsen_GB
dc.rightsArchived with thanks to Statistical Methods in Medical Researchen_GB
dc.titleAn alternative classification to mixture modeling for longitudinal counts or binary measuresen
dc.identifier.journalStatistical Methods in Medical Researchen_GB

Related articles on PubMed

All Items in MSF are protected by copyright, with all rights reserved, unless otherwise indicated.