On zero-distorted generalized geometric distribution

ABSTRACT

We propose a new generalized geometric distribution which permits inflation/deflation of the zero count probability and study some of its properties. We also present an actuarial application of this distribution and fit it to three datasets used by other researchers. It is observed that the proposed distribution fits reasonably well to these data. Further, in a regression setup, the performance of this distribution is studied vis–a–vis other competing distributions used for explaining variability in a response variable.

Keywords:

• Geometric distribution
• Exponential distribution
• Maximum likelihood function (MLE)
• Mixture models
• Zero-distortion parameter
• Zero-distorted regression

Mathematics Subject Classification:

• 60E05, 62E99

Acknowledgment

The authors are thankful to Prof. H.J. Vaman for his comments.