Risk performance of a pre-test ridge regression estimator under the LINEX loss function when each individual regression coefficient is estimated

摘要

In this paper, we consider a linear regression model and propose a pre-test ridge regression estimator which is obtained by incorporating a pre-test into a ridge regression estimator proposed by Huang [J.-C. Huang, Improving the estimation precision for a selected parameter in multiple regression analysis: An algebraic approach, Econ. Lett. 62 (1999), pp. 261-264]. We derive the exact formula for the risk of the estimator under the asymmetric LINEX loss function. Our numerical results show that the risk performance of the estimator is improved by incorporating the pre-test.