An alternative approach to the Kalman's scheme of identification in the presence of multicollinearity

Abstract

In this paper we present an alternative approach to the problem of multicollinearity stemming from Kalman's system theoretic approach on identification from noisy data, and link it to the most frequently used in practice, when dealing with multicollinear data, procedure of ridge regression. Especially we describe Kalman's scheme of identification and discuss its use in the presence of multicollinearity and its association with ridge regression.