No label defined (Q4373563)

Consider the distribution function \(H(x)=\int\limits_\Theta G(x|\theta) dF(\theta)\) of a mixture of \(G\) and \(F\), where \(G\) is known and \(F\) is unknown. The problem is to estimate the expectation of \(\gamma(\theta)\), based on a sample \((X_1,\dots,X_n)\) for \(H\). A general method is given for deriving an unbiased estimator. The method is demonstrated using the Weibull, noncentral \(t\), hypergeometric, normal, binomial and negative binomial distributions. It is shown how the method may be applied to certain estimation problems occurring in statistical quality control. For the entire collection see [Zbl 0867.00026].