|
Activity Number:
|
475
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 5, 2009 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
IMS
|
| Abstract - #304843 |
|
Title:
|
A Goodness-of-Fit Test for a Generalized Inverse Gaussian Based on the Matsumoto-Yor Property
|
|
Author(s):
|
Severien Nkurunziza*+ and Angelo E. Koudou
|
|
Companies:
|
University of Windsor and Institut Elie Cartan
|
|
Address:
|
401 Sunset Av., Windsor, ON, N9B3P4, Canada
|
|
Keywords:
|
Gamma distribution ; generalized inverse Gaussian distribution ; Matsumoto-Yor property ; goodness-of-fit test
|
|
Abstract:
|
The generalized inverse Gaussian has been revealed to be more convenient in modeling and analyzing the data sets where the observations are positive and highly right-skewed. In this communication, we consider to derive a goodness-of-fit test for testing if a given data set is from a generalized inverse Gaussian distribution. The established goodness-of-fit test is based on the extended version of the Matsumoto-Yor property. Further, in order to illustrate the application of the suggested test, we analyze four real data sets. Finally, numerical results from intensive simulation studies show that the suggested test is consistent.
|
