Some Robust Liu Estimators

Authors

  • Adewale Lukman
  • Kayode Ayinde 1Department of Statistics, Ladoke Akintola University of Technology, P.M.B. 4000, Ogbomoso, Oyo State, Nigeria
  • Ajiboye S. Adegoke Department of Statistics, Federal University of Technology, Akure
  • Daramola Tosin Department of Statistics, Ladoke Akintola University of Technology, P.M.B. 4000, Ogbomoso, Oyo State, Nigeria

Keywords:

Ordinary Least Square Estimator, Liu Estimator, Robust Estimators

Abstract

In a classical linear regression model, Liu and Robust Estimators were developed to deal with the problem of multicollinearity and
outliers respectively. This paper proposes some robust Liu estimators (RLEs) to jointly address the problem of multicollinearity and
outliers and illustrates the proposed estimators with real life data sets. Based on the performances of these estimators using the
Mean Square Error criterion, results show that the Robust Liu Estimators perform better than the ordinary least square (OLS), Liu
estimator and Robust estimators when data sets suffer both problems. Furthermore, it is observed that the M Robust Liu Estimator
(MRLE) is most efficient when outliers are in the y-direction; and when outliers are in the x or both y and x direction, the LTS Robust
Liu Estimator (LTSRLE) is most efficient

Downloads

Download data is not yet available.

Author Biography

Adewale Lukman

In a classical linear regression model, Liu and Robust Estimators were developed to deal with the problem of multicollinearity and outliers respectively. This paper proposes some robust Liu estimators (RLEs) to jointly address the problem of multicollinearity and outliers and illustrates the proposed estimators with real life data sets. Based on the performances of these estimators using the Mean Square Error criterion, results show that the Robust Liu Estimators perform better than the ordinary least square (OLS), Liu estimator and Robust estimators when data sets suffer both problems. Furthermore, it is observed that the M Robust Liu Estimator (MRLE) is most efficient when outliers are in the y-direction; and when outliers are in the x or both y and x direction, the LTS Robust Liu Estimator (LTSRLE) is most efficient.

Downloads

Published

2017-04-10 — Updated on 2017-04-10

Versions

How to Cite

Lukman, A., Ayinde, K., Adegoke, A. S., & Tosin, D. . (2017). Some Robust Liu Estimators. Zimbabwe Journal of Science and Technology, 12(1), 8–14. Retrieved from https://journals.nust.ac.zw/index.php/zjst/article/view/101

Issue

Vol. 12 No. 1 (2017): Volume Twelve

Section

Articles