A generalization of the univariate slash by a scale-mixtured exponential power distribution

Genc A. İ.

COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION, vol.36, no.5, pp.937-947, 2007 (SCI-Expanded) identifier identifier

  • Publication Type: Article / Article
  • Volume: 36 Issue: 5
  • Publication Date: 2007
  • Doi Number: 10.1080/03610910701539161
  • Journal Name: COMMUNICATIONS IN STATISTICS-SIMULATION AND COMPUTATION
  • Journal Indexes: Science Citation Index Expanded (SCI-EXPANDED), Scopus
  • Page Numbers: pp.937-947
  • Keywords: exponential power distribution, heavy tail, M-estimation, robustness, slash distribution, LOCATION PARAMETER, ROBUST, REGRESSION
  • Çukurova University Affiliated: No

Abstract

A generalization of the slash distribution is derived using the scale mixture of the exponential power distribution. The newly defined family of distributions provides a rich flexibility on the tail heaviness and yields alternative robust estimators of location and scale in non normal situations. In order to investigate asymptotically the bias properties of the estimators, a simulation study is performed. The performance of the estimators on two well-known real data sets is also illustrated.