In this paper, we introduce the generalized jackknifed ridge (GJR) estimator and  jackknifed ridge  estimator in a linear model with correlated or heteroscedastic errors. Therefore, we extend the jackknifed ridge (JR) estimator proposed by Ozkale [1]. Under a mean square error (MSE) and some conditions, the GJR estimator is superior to the generalized ridge (GR) estimator, and  estimator is superior to the GJR and  ridge  estimators. Furthermore, we build several models for GPS data of dam deformation by these estimation methods, and the application is given to illuminate our theoretical results.