Kibria-Lukman Estimator for General Linear Regression Model with AR(2) Errors: A Comparative Study with Monte Carlo Simulation

The sensitivity of the least-squares estimation in a regression model is impacted by multicollinearity and autocorrelation problems. To deal with multicollinearity, Ridge, Liu, Ridge-type biased estimators have been presented statistical literature. recently proposed Kibria-Lukman estimator one estimators. literature has compared others using mean square error criterion for linear model. It was achieved study conducted on estimator's performance under first-order autoregressive erroneous autocorrelation. When there an problem second-order, evaluating according to makes this paper original. scalar second-order structure evaluated Monte Carlo simulation two real examples, Generalized Least-squares, Liu estimators.
 findings revealed that when variance small, gave very close values popular As grew, did not give fairly similar as small variance. However, outperformed Least-Squares all possible cases.