Robust designs for models with possible bias and correlated errors

(整期优先)网络出版时间:2010-03-13
/ 1
Thispaperstudiesthemodel-robustdesignproblemforgeneralmodelswithanunknownbiasorcontaminationandthecorrelatederrors.ThetrueresponsefunctionisassumedtobefromareproducingkernelHilbertspaceandtheerrorsarefittedbytheqthordermovingaverageprocessMA(q),especiallytheMA(1)errorsandtheMA(2)errors.Inbothsituations,designcriteriaarederivedintermsoftheaverageexpectedquadraticlossfortheleastsquaresestimationbyusingaminimaxmethod.Acaseisstudiedandtheorthogonalityofthecriteriaisprovedforthisspecialresponse.Therobustnessofthedesigncriteriaisdiscussedthroughseveralnumericalexamples.