A fast approximate fitting for mixture of multivariate skew t-distribution via EM algorithm

A mixture of multivariate canonical fundamental skew t -distribution (CFUST) has been of interest in various fields. In particular, interest in the unsupervised learning society is noteworthy. However, fitting the model via EM algorithm su ff ers from significant processing time. The main cause is due to the calculation of many multivariate t -cdfs (cumulative distribution functions) in E-step. In this article, we provide an approximate, but fast calculation method for the in univariate fashion, which is the product of successively conditional univariate t -cdfs with Taylor’s first order approximation. By replacing all multivariate t -cdfs in E-step with the proposed approximate versions, we obtain the admissible results of fitting the model, where it gives 85% reduction time for the 5 dimensional skewness case of the Australian Institution Sport data set. For this approach, discussions about rough properties, advantages and limits are also presented.

• Publication year

  2020

• Venue

  Communications for Statistical Applications and Methods

• Publication date

  2020-03-31

• Fields of study

  Mathematics, Computer Science

• Identifiers
  DOI 10.29220/csam.2020.27.2.255
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Robust mixtures of factor analysis models using the restricted multivariate skew-t distribution

  2018influential reference
• An approximate fitting for mixture of multivariate skew normal distribution via EM algorithm

  2016cited by this paper
• A Simple Parallel EM Algorithm for Statistical Learning via Mixture Models

  2016cited by this paper
• Finite mixtures of canonical fundamental skew t -distributions The unification of the restricted and unrestricted skew t -mixture models

  2015cited by this paper
• Approximations to the t distribution

  2015cited by this paper
• Maximum Likelihood Estimation for Finite Mixtures of Canonical Fundamental Skew t-Distributions: the Unification of the Unrestricted and Restricted Skew t-Mixture Models

  2014cited by this paper
• Finite mixtures of canonical fundamental skew $$t$$t-distributions

  2014cited by this paper
• Some results on the truncated multivariate t distribution

  2012influential reference
• Robust mixture modeling using multivariate skew t distributions

  2010cited by this paper
• Automated high-dimensional flow cytometric data analysis

  2009influential reference
• Computation of Multivariate Normal and t Probabilities

  2009cited by this paper
• On fundamental skew distributions

  2005influential reference
• An Introduction to Regression Graphics

  1995cited by this paper
• An Introduction to Regression Graphics

  1994cited by this paper

• No citing papers are available for this paper.