Poisson-modification of the Quasi Lindley optimal generalized regression estimator

Abstract

In this paper, we propose an optimal generalized regression estimator based on the sample information for the Poisson-Modification of the Quasi Lindley (PMQL) regression model namely the PMQL optimal generalized regression estimator. Asymptotic properties of the proposed estimator are obtained. Further, the conditions of superiority over some other regression coefficient estimators for the PMQL regression model are also derived based on the mean square error criterion. In order to compare the performance of the proposed estimator over some other PMQL regression coefficient estimators in the scalar mean square error sense, a Monte Carlo simulation study is designed and results are discussed. Finally, a simulated data example and a real-world example are taken to show the potentiality of the new regression model based on the proposed estimator and show the consistency of the theoretical and simulation results, respectively.

Keywords:

• PMQL distribution
• PMQL regression
• regression coefficient estimators
• multicollinearity
• biased estimators

Acknowledgments

The authors would like to thank the Postgraduate Institute of Science, University of Peradeniya, Sri Lanka for providing all facilities to do this research and the editor-in-chief and the referees for their helpful comments, which significantly improved the paper.

Disclosure statement

No potential conflict of interest was reported by the author(s).