A COMPARATIVE STUDY OF REMEDIES OF THE MULTICOLLINEARITY PROBLEM IN THE POISSON REGRESSION MODEL WITH APPLICATION

The Poisson regression model is considered as one of the most popular and used generalized linear models while dealing with numerical data, that is when the response variable is expressed as integer numbers, which are estimated using the maximum likelihood estimator. This data is vulnerable to the occurrence of the multiple linear correlation problem. It provides errors in the result of the maximum possibility estimator and inflation in the value of the variance. To treat this problem, many estimators have been proposed to give better results than the maximum possibility estimator. In this article, we propose a new estimator and compare it with previously known estimators. The quality of the proposed estimator has been proven through an application to realistic data using Scalar Mean Square Error (SMSE). Received: August 7, 2024 Revised: August 21, 2024 Accepted: September 3, 2024