Advances and Applications in Statistics
Volume 10, Issue 2, Pages 179 - 193
(December 2008)
|
|
BAYESIAN ESTIMATION OF PARAMETERS IN SOME BIVARIATE EXPONENTIAL MODELS
David D. Hanagal (India) and Kambiz Ahmadi Angali (India)
|
Abstract: In this paper, we obtain Bayesian estimate of parameters for bivariate exponential of Marshal-Olkin [14] when the components are in series and parallel. We use the conjugate prior and its parameters are estimated by consistent estimators, in addition we extend this method for the bivariate exponential (BVE) of Proschan-Sullo [16]. Simulation study has been conducted to validate the proposed method and we show that the estimated values are close to true values. Also we show that Bayesian analysis is easy to handle and choosing an appropriate prior leads to the accurate and precise estimation of parameters of BVED. An eye clinic data is used, which fits BVE of Proschan-Sullo [16]. The Bayesian estimation and MLE are compared using this data and the results show that Bayesian estimated values are close to MLE’s with low standard error (SE). |
Keywords and phrases: Bayesian estimation, bivariate exponential, conjugate prior, simulation study. |
|
Number of Downloads: 333 | Number of Views: 1226 |
|