Maximum likelihood parameter estimation for the multivariate skew-slash distribution

ARSLAN O.

STATISTICS & PROBABILITY LETTERS, vol.79, no.20, pp.2158-2165, 2009 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 79 Issue: 20
  • Publication Date: 2009
  • Doi Number: 10.1016/j.spl.2009.07.009
  • Journal Name: STATISTICS & PROBABILITY LETTERS
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.2158-2165
  • Ankara University Affiliated: No

Abstract

In this paper we consider the parameter estimation of the multivariate skew-slash distribution introduced by Arslan [Arslan, O. 2008. An alternative multivariate skew-slash distribution. Statistics and Probability Letters 78, 2756-2761], which is a recent example of the normal variance-mean mixture distributions. Due to the complexity of the likelihood function, estimation of its parameters by direct maximization of the likelihood function seems difficult. To overcome this problem, we propose a simple EM-based maximum likelihood estimation procedure to estimate the parameters of the multivariate skew-slash distribution. We provide three examples to demonstrate the modeling strength of the multivariate skew-slash distribution and the feasibility of the proposed EM algorithm. (C) 2009 Elsevier B.V. All rights reserved.