Estimator of the Parameter in a Subclass of the Exponential Model under Weighted p,q Symmetric Entropy Loss

For exponential model c(x,n)θ(~-v)e~(-T(x))/θ,we dealt with the Bayes estimator and admissible estimator of scale parameterθunder a new loss-weighted p,q symmetric entropy loss L(θ,d) =θ~p/pδ~p+δ~q/qθ~q -2(p,q 0) which deduced from entropy that related to theory of information.The general and exact form of the Bayes estimator were obtained. The admissibility and inadmissibility of a class of linear estimators of the form cT(X) + d were studied.The invariance of the two estimators was proved using the theorem integral transform.