Abstract
The problem of multilinearity that occurs as a result of the high correlation and linear correlation between two or more explanatory variables included in its construction is one of the most important model building problems, which negatively affects the process of estimating the regression model parameters. This study aims to review the shrinking estimators that are used to address the problem of multicollinearity that appears in counting data models, especially the Poisson regression model. This model is considered one of the most used models in the case that the data of the response variable have countable values and are not subject to a normal distribution. Through Monte-Carlo simulation experiments, it was found that the two-parameter shrink estimator is the best proposed estimator because it reduces the mean square error of the model.