Author(s): |
Ping Wu, Department of the Basics, Air Force Radar Academy, Wuhan, 430019, China Qingguo Liu, Department of the Basics, Air Force Radar Academy, Wuhan, 430019, China Siru Zhu, Department of the Basics, Air Force Radar Academy, Wuhan, 430019, China |
Abstract: |
Focus on the shortage of the multicollinearity in ordinary restricted least square estimation with parameter estimate. In order to overcome this problem, based on the stochastic linear restrictions, a new estimation as restricted linear k-d-type estimation is proposed. In the mean squared error sense, discuss its properties, draw four main results, compare with the ordinary restricted least squares estimation ,the restricted ridge estimation and the restricted Liu estimation, the method we proposed was superior.
|