A new robust ridge parameter estimator having no outlier and ensuring normality for linear regression model

• DOI: 10.1016/j.jrras.2023.100788
• Publication Date: 2023-12-14T12:47:56Z
• Authors: Selman Mermi; Özge Akkuş; Atila Göktaş; Necla Gündüz
• Keywords: Robust regression; Variance Inflation Factor

In order to accurately estimate the regression coefficients in a multiple linear model having multicollinearity, ridge is well-liked biased estimation technique used as an alternative least squares method. So far, numerous estimators have been proposed for parameter this study, robust estimator that performs better than 366 different literature date has developed using "search method". contrast studies literature, addition Mean Square Error (MSE) criterion both and assessed terms of outliers normality. Furthermore, simulation design with total 6050 cases were included study which corresponds values varied widely number independent variables (p), sample size (n), correlation coefficient between (ρ), standard deviation errors (σ). Significant inter associations among MSE, outlier detection normality criteria found. Results from conducted revealed was most effective one these three criteria, regardless regressors, size, multicollinearity level, errors.