Modified ridge-type estimator for the zero inflated negative binomial regression model

Abstract

The Zero-inflated negative binomial (ZINB) regression models are commonly used for count data that shows an over-dispersion and extra zeros. Multicollinearity is considered to be a significant issue in the estimation of parameters in the ZINB regression model. Thus, to alleviate the negative effects of multicollinearity, a new estimator called ZINB modified ridge type (ZINBMRT) estimator is proposed. Furthermore, we proposed some new approaches to estimate the shrinkage parameters for the ZINBMRT estimator. A Monte Carlo simulation study and illustrative example are given to show the superiority of the proposed ZINBMRT estimator over some of the existing estimation methods. Based on the findings of simulation study and example, it is observed that the proposed ZINBMRT estimator under different suggested parameters give a better performance over the other competitive estimators.