PENDUGAAN PARAMETER REGRESI ROBUST METODE MINIMUM COVARIANCE DETERMINANT DAN METODE TELBS
Abstract
The parameter estimator on the regression model can be obtained through the ordinary least square (OLS). When there are outliers in the data, OLS cannot be applied because it will produce an unbiased estimator that is not the best linear estimator. Another alternative to addressing the presence of outlier data without deleting the data is robust regression. Robust regression methods include the minimum covariance determinant (MCD) and the TELBS method. This study aims to determine the estimation of regression parameters produced using the MCD and TELBS methods when entering outlier data. The data used are simulation data with various levels of outliers, namely 5%, 10%, and 20%. The outliers inserted are the outliers on variable X, variable Y, and variables X and Y. The result of this study is that the robust regression methods of MCD and TELBS both produce unbiased parameter estimators when there are outlier data.
Downloads
This work is licensed under a Creative Commons Attribution 4.0 International License.
This work is licensed under a Creative Commons Attribution 4.0 International License.