The Use of Empirical Bayes Techniques in Testing Parameteric Hypotheses | ||||
The Egyptian Statistical Journal | ||||
Article 2, Volume 25, Issue 1, June 1981, Page 7-23 PDF (9.8 MB) | ||||
Document Type: Original Article | ||||
DOI: 10.21608/esju.1981.428122 | ||||
![]() | ||||
Authors | ||||
E. Al-Hussaini* ; M.A. Mohamad Ali | ||||
Dep. of Mathematics, University of Assiut, Assiut, Egypt | ||||
Abstract | ||||
Empirical Bayes (EB) rules are obtained for testing the following parametric cases: k-simple and two composite (one and two-sided) hypotheses. A step-function type prior distribu-tion, with unknown jumps, is a suggested form in place of the unknown form prior, used in testing k-simple hypotheses. The prior distribution, in the case of testing composite hypotheses, is not assumed known and no attempt is made to estimate it. Estimates, based on previous random samples, of marginal pro-bability functions are rather used in setting up the EB decisions. | ||||
Statistics Article View: 66 PDF Download: 12 |
||||