Affiliation:
1. Information and Engineering University
Abstract
In this paper, the GM estimation is integrated with ridge estimation, principal component estimation and LIU estimation, resulting in a new class of robust unbiased estimation, and given the appropriate method of calculating. The example shows that such a new biased estimate is not only resistant to the interference of design matrix multicollinearity, but also withstands the adverse effects of outliers and high leverage points. They are really better than the LS estimation, unbiased estimation, robust M estimation and robust M-type biased estimation.
Publisher
Trans Tech Publications, Ltd.
Reference22 articles.
1. Gruber, M.H.J., Improving Efficiency by Shrinkage: The James Stein and Ridge Regression Estimators. Marcel Dekker, Inc.: New York, (1998).
2. Hoer, A. E., and Kennard, R. W., Ridge Regression: Biased Estimation for Non-orthogonal Problems, Techno metrics, 12, pp.55-67, 1970a.
3. Hoer, A. E., and Kennard, R. W., Ridge Regression: Applications to Non-orthogonal Problems, Techno metrics, 12, pp.69-82, 1970b.
4. Massy, W.F., Principal components regression in exploratory statistical research., JASA, 60, pp.234-266, (1965).
5. Liu, K, A new class of biased estimate in linear regression., Communications in Statistics – Theory and Methods, 22 (2), p.393–402, (1993).