AL-Qadisiyah Journal For Administrative and Economic sciences

A Comparative study of traditional model selection methods with some of the regularization methods

Document Type : Research Paper

Authors

Mohammed H. AL-Sharoot
Amal H. Alwan

University of AL-Qadisiya

https://doi.org/10.33916/qjae.2026.014556
Abstract
Linear regression models are used to describe and estimate the relationship between a response variable and a set of covariates. However, if some of covariates are inactive in the regression, the estimated relationship could be unstable and unpredictable. Many methods have been used over the years to identify the active covariates in the regression. In this paper, we propose Bayesian bridge-randomized expectile regression (BBRER). We compare the proposed method with the traditional model selection methods with Lasso and adaptive Lasso methods. Simulation methods show that all methods are perform comparably, however; Lasso performs the best in 80% of the simulation studies. Real data analyses using prostate cancer data also show that Lasso is the best.
Crossmark
AL-Qadisiyah Journal For Administrative and Economic sciences
Volume 28, Issue 1
Spring 2026
Pages 45-56

Home

Guide for Authors

Submit Manuscript

Reviewers

Contact Us