Limiting Form of the Bayes Estimate of the Regression Coefficients in the Presence of Nonstationary Noise

Sufficient conditions are obtained for the asymptotic invariance of the Bayes estimates of the regression coefficients relative to a fairly extensive class of a priori distributions in the presence of nonstationary normal noise. The limiting form of the Bayes estimate is deduced. Conditions are determined which ensure a sufficiently precise approximation of the Bayes estimate to its limiting form.