Abstract
We consider the problem of the choice of an estimator and an experimental design for estimating the response surface in a linear regression model according to Bayes-and minimax-Bayes-optimality. The well-known Bayes estimator w.r.t. normally i.i.d. observations, conjugate prior distribution and quadratic loss is shown to have a satisfying robustness with regard to the optimality criteria, if we change to larger classes of prior and error distributions (e.g. all distributions with fixed first and second moments) and more general loss functions (e.g. monotone and convex functions). Moreover, for this Bayes estimator we can point out a remarkable robustness of'designing against the loss function; e.g. minimax designs under quadratic loss are also minimax under any convex loss function.
1Paper presented to the poster session of the Fourth International Summer School on Problems of Model Choice and Parameter Estimation in Linear Regression, Mühlhausen (G.D.R.), May 1979
1Paper presented to the poster session of the Fourth International Summer School on Problems of Model Choice and Parameter Estimation in Linear Regression, Mühlhausen (G.D.R.), May 1979
Notes
1Paper presented to the poster session of the Fourth International Summer School on Problems of Model Choice and Parameter Estimation in Linear Regression, Mühlhausen (G.D.R.), May 1979