Maximum likelihood estimation for multivariate skew normal mixture models

A feasible EM algorithm is developed for finding the maximum likelihood estimates of parameters in this context and a general information-based method for obtaining the asymptotic covariance matrix of themaximum likelihood estimators is presented.

Abstract

This paper provides a flexible mixture modeling framework using the multivariate skew normal distribution. A feasible EM algorithm is developed for finding the maximum likelihood estimates of parameters in this context. A general information-based method for obtaining the asymptotic covariance matrix of the maximum likelihood estimators is also presented. The proposed methodology is illustrated with a real example and results are also compared with those obtained from fitting normal mixtures.