vix.ing · top · new · best · stats · spec

Maximum Likelihood Estimation for Finite Mixtures of Canonical\n Fundamental Skew t-Distributions: the Unification of the Unrestricted and\n Restricted Skew t-Mixture Models

2014/01/31 by Sharon Lee, Geoffrey J. McLachlan, Lee, Sharon X. +1
Computer Science · Mathematics · #Bayesian Methods and Mixture Models #FOS: Computer and information sciences #Methodology (stat.ME) #Statistical Distribution Estimation and Applications #Statistical Methods and Bayesian Inference

paper · pdf · doi:10.48550/arxiv.1401.8182

openalex publication_date 2014/01/31 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28

Abstract

In this paper, we present an algorithm for the fitting of a location-scale\nvariant of the canonical fundamental skew t (CFUST) distribution, a superclass\nof the restricted and unrestricted skew t-distributions. In recent years, a few\nversions of the multivariate skew t (MST) model have been put forward,\ntogether with various EM-type algorithms for parameter estimation. These\nformulations adopted either a restricted or unrestricted characterization for\ntheir MST densities.\n In this paper, we examine a natural generalization of these developments,\nemploying the CFUST distribution as the parametric family for the component\ndistributions, and point out that the restricted and unrestricted\ncharacterizations can be unified under this general formulation. We show that\nan exact implementation of the EM algorithm can be achieved for the CFUST\ndistribution and mixtures of this distribution, and present some new analytical\nresults for a conditional expectation involved in the E-step.\n

Related