Estimation of the Premium in the Contaminated Pareto Distribution and Detecting of Outliers
Seminar in Insurance and Economics
SPEAKER: Mehdi Jabbari Nooghabi (Ferdowsi University of Mashhad).
TITLE: Estimation of the Premium in the Contaminated Pareto Distribution and Detecting of Outliers.
ABSTRACT: This seminar is for presentation of the results of four research papers which are published or under the process. Here, I present estimation of the parameters of the contaminated Pareto model as well as the Premium, in different situations. Also, when there is some reason to be outliers in the data, the problem is that how to presence with outliers, either removing/ignoring or using nonparametric methods such as robust or detecting outliers by using parametric or nonparametric methods. We know that the parametric methods are always better than the others, but it is complicated or maybe very difficult to proof the theorems related. So, here we estimate the parameters of a Pareto distribution in the presence of outliers by using an outlier’s model and then estimate the Premium. The other approach is to detect outliers in the Pareto distribution based on the parametric methods. Finally, we present some actual example from insurance companies and see how the method are perform. To easier using of the mathematical methods, some applications are created and presented.