Penalized Maximum Likelihood Estimator for Skew Normal Mixtures

TL;DR

This paper introduces a penalized maximum likelihood approach for skew normal mixture models, addressing issues of unbounded likelihood and diverging shape parameters, with proven consistency and practical algorithms.

Contribution

It proposes a novel penalized likelihood method for skew normal mixtures, ensuring estimator consistency and providing algorithms for practical computation.

Findings

Penalized estimators are strongly consistent.

Algorithms effectively compute estimators in practice.

Method outperforms existing approaches in simulations and real data.

Abstract

Skew normal mixture models provide a more flexible framework than the popular normal mixtures for modelling heterogeneous data with asymmetric behaviors. Due to the unboundedness of likelihood function and the divergency of shape parameters, the maximum likelihood estimators of the parameters of interest are often not well defined, leading to dissatisfactory inferential process. We put forward a proposal to deal with these issues simultaneously in the context of penalizing the likelihood function. The resulting penalized maximum likelihood estimator is proved to be strongly consistent when the putative order of mixture is equal to or larger than the true one. We also provide penalized EM-type algorithms to compute penalized estimators. Finite sample performances are examined by simulations and real data applications and the comparison to the existing methods.