On the estimation of ridge penalty in linear regression: Simulation and application

IF 1.2 4区综合性期刊 Q3 MULTIDISCIPLINARY SCIENCES

Kuwait Journal of Science Pub Date : 2024-06-06 DOI:10.1016/j.kjs.2024.100273

Muhammad Shakir Khan , Amjad Ali , Muhammad Suhail , Eid Sadun Alotaibi , Nahaa Eid Alsubaie

{"title":"On the estimation of ridge penalty in linear regression: Simulation and application","authors":"Muhammad Shakir Khan , Amjad Ali , Muhammad Suhail , Eid Sadun Alotaibi , Nahaa Eid Alsubaie","doi":"10.1016/j.kjs.2024.100273","DOIUrl":null,"url":null,"abstract":"<div><p>According to existing literature, the ordinary least squares (OLS) estimators are not the best in presence of multicollinearity. The inability of OLS estimators against multicollinearity has paved the way for the development of various ridge type estimators for circumventing the problem of multicollinearity. In this paper improved two-parameter ridge (ITPR) estimators are proposed. A simulation study is used to evaluate the performance of proposed estimators based on minimum mean squared error (MSE) criterion. The simulative results reveal that, based on minimum MSE, ITPR2 was the most efficient estimator compared to the considered estimators in the study. Finally, a real-life dataset is analyzed to demonstrate the applications of the proposed estimators and also checked their efficacy for mitigation of multicollinearity.</p></div>","PeriodicalId":17848,"journal":{"name":"Kuwait Journal of Science","volume":"51 4","pages":"Article 100273"},"PeriodicalIF":1.2000,"publicationDate":"2024-06-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.sciencedirect.com/science/article/pii/S2307410824000981/pdfft?md5=c29d7ee62440e157955f85400deb96fb&pid=1-s2.0-S2307410824000981-main.pdf","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"Kuwait Journal of Science","FirstCategoryId":"103","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S2307410824000981","RegionNum":4,"RegionCategory":"综合性期刊","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q3","JCRName":"MULTIDISCIPLINARY SCIENCES","Score":null,"Total":0}

引用次数: 0

Abstract

According to existing literature, the ordinary least squares (OLS) estimators are not the best in presence of multicollinearity. The inability of OLS estimators against multicollinearity has paved the way for the development of various ridge type estimators for circumventing the problem of multicollinearity. In this paper improved two-parameter ridge (ITPR) estimators are proposed. A simulation study is used to evaluate the performance of proposed estimators based on minimum mean squared error (MSE) criterion. The simulative results reveal that, based on minimum MSE, ITPR2 was the most efficient estimator compared to the considered estimators in the study. Finally, a real-life dataset is analyzed to demonstrate the applications of the proposed estimators and also checked their efficacy for mitigation of multicollinearity.

查看原文本刊更多论文

关于线性回归中脊惩罚的估计：模拟与应用

根据现有文献，在存在多重共线性的情况下，普通最小二乘法（OLS）估计器并非最佳估计器。OLS 估计器无法解决多重共线性问题，这为开发各种脊型估计器来规避多重共线性问题铺平了道路。本文提出了改进的双参数脊（ITPR）估计器。模拟研究根据最小均方误差（MSE）标准来评估所提出的估计器的性能。模拟结果表明，根据最小均方误差标准，ITPR2 与研究中考虑的估计器相比是最有效的估计器。最后，分析了一个真实数据集，以展示所提出的估计器的应用，并检验其在缓解多重共线性方面的功效。

本文章由计算机程序翻译，如有差异，请以英文原文为准。

求助全文

约1分钟内获得全文求助全文

来源期刊

Kuwait Journal of Science MULTIDISCIPLINARY SCIENCES-

CiteScore

1.60

自引率

28.60%

发文量

132

期刊介绍： Kuwait Journal of Science (KJS) is indexed and abstracted by major publishing houses such as Chemical Abstract, Science Citation Index, Current contents, Mathematics Abstract, Micribiological Abstracts etc. KJS publishes peer-review articles in various fields of Science including Mathematics, Computer Science, Physics, Statistics, Biology, Chemistry and Earth & Environmental Sciences. In addition, it also aims to bring the results of scientific research carried out under a variety of intellectual traditions and organizations to the attention of specialized scholarly readership. As such, the publisher expects the submission of original manuscripts which contain analysis and solutions about important theoretical, empirical and normative issues.