Choosing ridge Parameters in the Linear regression Model with AR(1) Error: A Comparative Simulation Study

Abstract

The performance of shrinkage ridge estimators in the linear regression model has been studied to reduce the effect of multicolinearity on the estimation of the Imur regresston parameters, Trinkler (1984) proposed a ridge estimator in the linear regresston model when the assump tton of uncorrelatedness ts not sartafled. Since there is no attempt to study the recent types of estimated ridge parumeter when the assumption of uncorrelataincar u not satuled, thu paper tries to show the performance of some ridge rasimators in the linear regression model with correlatal error based on the minimum man squared error (MSE) eruerton. A si fatton study and a summersoal example have been made to esaluate the performance of these attrators of edge pararserk The mulatton study supports that sums ridge attmators are prembing and can be recommend for the practitionery