We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
Jackknifing Ridge Estimator for Logistic Regression Model.
- Authors
Hammood, Nawal Mahmood; Algamal, Zakariya Yahya
- Abstract
It has been repeatedly shown that the ridge regression model is a desirable shrinking technique to lessen the consequences of multicollinearity. When the response variable involves binary data, the logistic regression model is a well-known model in use. However, it is well known that multicollinearity has a detrimental impact on the variance of the logistic regression coefficients' maximum likelihood estimate. Numerous scholars have suggested a logistic ridge estimator as a solution to this issue. The jackknifing logistic ridge estimator (JLRE) is suggested and derived in this study. The goal of the JLRE is to obtain a diagonal matrix with low diagonal element values, which will reduce the shrinkage parameter and enable a better, less biassed estimator to be produced. According on the results of our Monte Carlo simulation, the JLRE estimator can significantly outperform other available estimators. Additionally, the JLRE estimator surpasses the logistic ridge estimator and the maximum likelihood estimator in terms of predictive performance, according to the practical application findings.
- Subjects
MULTICOLLINEARITY; LOGISTIC regression analysis; REGRESSION analysis; MONTE Carlo method; MAXIMUM likelihood statistics; POCKETKNIVES
- Publication
Pakistan Journal of Statistics & Operation Research, 2022, Vol 18, Issue 4, p955
- ISSN
1816-2711
- Publication type
Article
- DOI
10.18187/pjsor.v18i4.3748