RT info:eu-repo/semantics/article
T1 Improving model choice in classification: an approach based on clustering of covariance matrices
A1 Rodríguez Vítores, David
A1 Matrán Bea, Carlos
K1 Parsimonious model
K1 Gaussian mixture model
K1 Bayesian information criterion
K1 Model-based classification
K1 EM algorithm
K1 12 Matemáticas
AB This work introduces a refinement of the Parsimonious Model for fitting a Gaussian Mixture. The improvement is based onthe consideration of clusters of the involved covariance matrices according to a criterion, such as sharing Principal Directions.This and other similarity criteria that arise from the spectral decomposition of a matrix are the bases of the ParsimoniousModel. We show that such groupings of covariance matrices can be achieved through simple modifications of the CEM(Classification Expectation Maximization) algorithm. Our approach leads to propose Gaussian Mixture Models for model-based clustering and discriminant analysis, in which covariance matrices are clustered according to a parsimonious criterion,creating intermediate steps between the fourteen widely known parsimonious models. The added versatility not only allowsus to obtain models with fewer parameters for fitting the data, but also provides greater interpretability. We show its usefulnessfor model-based clustering and discriminant analysis, providing algorithms to find approximate solutions verifying suitablesize, shape and orientation constraints, and applying them to both simulation and real data examples.
PB Springer
SN 0960-3174
YR 2024
FD 2024
LK https://uvadoc.uva.es/handle/10324/75137
UL https://uvadoc.uva.es/handle/10324/75137
LA eng
NO Statistics and Computing, 2024, vol. 34, n. 4
NO Producción Científica
DS UVaDOC
RD 21-abr-2025