URV's Author/s: Bhati, Deepesh Chakraborty, Subrata

Keywords: Aggregate claim, count regression, geometric distribution, transmuted distribution

Abstract: A two-parameter transmuted geometric distribution is proposed as a new generalization of the geometric distribution by employing the quadratic transmutation techniques of Shaw and Buckley. The additional parameter plays the role of controlling the tail length. Distributional properties of the proposed distribution are investigated. Maximum likelihood estimation method is discussed along with some data fitting experiments to show its advantages over some existing distributions in literature. The tail flexibility of density of aggregate loss random variable assuming the proposed distribution as primary distribution is outlined and presented along with a illustrative modelling of aggregate claim of a vehicle insurance data. Finally, we present a count regression model based on the proposed distribution and carry out its comparison with some established models.

Journal publication year: 2016

Publication Type: info:eu-repo/semantics/publishedVersion info:eu-repo/semantics/article