Sensitivity of ridge-type estimation methods to observation accuracy and sampling rate

Abstract

Ridge-type estimation methods have been shown to provide increased solution accuracy over minimum variance methods in ill-conditioned estimation problems. The accuracy gained by the use of such methods is a function of the level of ill conditioning present in the problem, which is dependent on the accuracy of the observations and the sampling rate (over a given observation span). This study presents results of a simulation designed to demonstrate the sensitivity of ridge-type estimation methods to the observation accuracy and the sampling rate used in the estimation process. The results are analyzed, and conclusions are drawn concerning the effectiveness of ridge-type estimation methods in the type of ill-conditioned problems addressed here.