On some weighted biasing parameters in ridge regression model

The optimal choice of the biasing parameter in ridge regression always plays an important role in the bias-variance trade-off. This paper analyses recent well-known estimators of (Qasim M, Månsson K, Amin M, et al. Biased adjusted Poisson’s ridge estimators: method and application. Iran J Sci Technol Trans A Sci. 2020;44(6):1775–1789; Suhail M, Chand S, Kibria BMG. Quantile-based estimation of biasing parameters in ridge regression model. Commun Stat - Simul Comput. 2020;49(10):2732–2744) and (Khalid N, Khan DM, Suhail M, et al. Exploring new estimators in ridge regression: addressing multicollinearity in economic and petroleum product data analysis. Kuwait J Sci. 2025;52(4):100448) and proposes weight-based biasing parameters. An extensive simulation study shows the best performance of the new estimators on the basis of mean squared error evaluation metrics. The new estimators outperform under both normal and some non-normal error distributions. The best performance of estimators for all the considered scenarios is summarized in the summary table. Three numerical examples further illustrate the simulation results. We recommend the use of new estimators in regression modelling when multicollinearity is the problem of concern.