Bivariate regression models based on compound Poisson distribution

ABSTRACT

Based on a compound Poisson distribution, new bivariate regression models are introduced and studied. The parameters of the bivariate regression models are estimated by using the maximum likelihood method. Two applications on real datasets are presented to illustrate the models. The results show that these models are compatible to other bivariate Poisson models.

KEYWORDS:

• Bivariate distribution
• Compound distribution
• Correlated count data
• Maximum likelihood estimation
• Regression models
• Zero-inflated models.

MATHEMATICS SUBJECT CLASSIFICATION:

• 62E15
• 62F10
• 62J02
• 62P10

Acknowledgments

The authors are grateful to the anonymous referees for their valuable suggestions that improved the presentation.

Additional information

Funding

This project was funded by the National Plan for Science, Technology and Innovation (MAARIFAH), King Abdulaziz City for Science and Technology, Kingdom of Saudi Arabia, Award Number (11-MAT1856-02).