NEW IMPROVED INFORMATIVE PRIORS FOR VARIANCE OF NORMAL DISTRIBUTION

by ABDUL HAQ .

Abstract: In this study a new informative prior is developed for unknown variance of the Normal distribution. The idea is to combine both informative and non-informative priors just as posterior distribution which combines likelihood and prior distribution, in order to update the information about unknown parameter of the probability distribution under study. Different new priors have been assumed for unknown variance of the Normal distribution. The comparison is based upon posterior variances, posterior coefficient of variability and posterior predictive variances. Finally, an improved informative prior for unknown Normal variance is suggested, which performs better than the priors given by Aslam and Haq (2009).

Key Words: Prior distribution, Density kernel, Posterior distribution, Posterior Predictive distribution

Author:
Abdul Haq, aaabdulhaq@yahoo.com

Editor: DEY, SANKU, sanku_dey2k2003@yahoo.co.in

READING THE ARTICLE: You can read the article in portable document (.pdf) format (224000 bytes.)

NOTE: The content of this article is the intellectual property of the authors, who retains all rights to future publication.

This page has been accessed 1911 times since NOVEMBER 25, 2009.


Return to the InterStat Home Page.