BAYES ESTIMATORS OF EXPONENTIAL PARAMETERS UTILIZING A GUESSED GUARANTEE WITH SPECIFIED CONFIDENCE

This paper proposes a procedure to modify the conjugate prior for the parameters of an exponential distribution in the light of an available guessed guarantee mu(0) with specified confidence c (0 < c < 1). The Bayes estimators under modified priors have been obtained. The proposed estimators have been compared with Bayes estimators under conjugate prior, shrinkage estimators and maximum likelihood estimators on the basis of a simulation study.