RT info:eu-repo/semantics/article
T1 Robust estimation for mixtures of Gaussian factor analyzers, based on trimming and constraints
A1 García Escudero, Luis Ángel
A1 Gordaliza Ramos, Alfonso
A1 Greselin, Francesca
A1 Salvatore, Ingrassia
A1 Mayo Iscar, Agustín
K1 Análisis multivariante
AB Mixtures of Gaussian factors are powerful tools for modeling an unobservedheterogeneous population, offering - at the same time - dimension reductionand model-based clustering. Unfortunately, the high prevalence of spurioussolutions and the disturbing effects of outlying observations, along maximum likelihoodestimation, open serious issues. In this paper we consider restrictions forthe component covariances, to avoid spurious solutions, and trimming, to providerobustness against violations of normality assumptions of the underlying latent factors.A detailed AECM algorithm for this new approach is presented. Simulationresults and an application to the AIS dataset show the aim and effectiveness of theproposed methodology.
PB Universidad de Valladolid. Facultad de Medicina
YR 2015
FD 2015
LK http://uvadoc.uva.es/handle/10324/11618
UL http://uvadoc.uva.es/handle/10324/11618
LA eng
NO Arxiv, Marzo 2015, vol. 1. p.1-30
NO Producción Científica
DS UVaDOC
RD 28-jun-2024