teigen : An R Package for Model-Based Clustering and Classification via the Multivariate t Distribution

The teigen R package is introduced and utilized for model-based clustering and classification. The tEIGEN family of mixtures of multivariate t distributions is formed via an eigen-decomposition of the component covariance matrices and subsequent componentwise constraints. The teigen package implemen...

Full description

Saved in:
Bibliographic Details
Published in:Journal of statistical software 2018-02, Vol.83 (7), p.1-32
Main Authors: Andrews, Jeffrey L., Wickins, Jaymeson R., Boers, Nicholas M., McNicholas, Paul D.
Format: Article
Language:English
Subjects:
classification
cluster analysis
mixture models
model-based clustering
multivariate t~distributions
teigen
Citations: Items that cite this one
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:The teigen R package is introduced and utilized for model-based clustering and classification. The tEIGEN family of mixtures of multivariate t distributions is formed via an eigen-decomposition of the component covariance matrices and subsequent componentwise constraints. The teigen package implements all previously published tEIGEN family members as well as eight additional models: four multivariate and four univariate. The resulting family of 32 mixture models is implemented in both serial and parallel, with useful dedicated functions. Methodology and examples that illustrate teigen's functionality are presented.
ISSN:1548-7660
1548-7660
DOI:10.18637/jss.v083.i07