University of Nigeria
Date & Time
October 14, 2022, 11:00 am – 12:00 pm
Topic:Modified jackknife kibria-lukman
estimator for the Poisson regression model
Abstract: Poisson regression is one of the methods to analyze count data and, the regression parameters are usually estimated using the maximum likelihood (ML) method. However, the ML method is sensitive to multicollinearity. Multicollinearity occurs when there is linear dependency among the explanatory variables and it often leads to unstable maximum likelihood estimates. We present a modified jackknife Poisson Kibria–Lukman (MJPKL) estimator to mitigate multicollinearity in the Poisson regression model. This estimator is theoretically compared with some existing estimators and the condition for its superiority is obtained. Simulation study and real-life applications were conducted to compare the performance of the estimators. The result shows that the MJPKLE gives better results than other estimators under some conditions. It also shows that the MJPKL estimator reduces the bias of the PKL estimator and dominates every estimator considered.