Year: 2025
Type: article
Abstract: In this study, we introduce three new ridge estimators, referred to as AIRE1, AIRE2, and AIRE3, designed to effectively address the issue of multicollinearity in regression analysis. The performance of these estimators was compared with the Ordinary Least Squares (OLS) and existing ridge estimators using the Mean Squared Error (MSE) as a benchmark. Theoretical comparisons of estimators are not possible. We used Monte Carlo simulations to assess their performance based on MSE. Although no single estimator works best across all scenarios of multicollinearity, the findings from simulations and real-world datasets demonstrated that the proposed estimators outperformed OLS and other ridge estimators. Among all the estimators, the newly proposed AIRE1 consistently outperformed the rest, establishing itself as a reliable and effective solution for mitigating multicollinearity.
Source: AIP Advances
DOI: 10.1063/5.0259923
Authors: Nadeem Akhtar, Muteb Faraj Alharthi
Institutions: Government of Khyber Pakhtunkhwa, Taif University
Language: English
Subfield: Statistics and Probability
Field: Mathematics
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $1,350
Funder: Taif University
OpenAlex ID: W4408612662