A Comparative study for estimation of the parameters of the folded exponential power Ddstribution

Folded distributions are commonly used for the data set which is obtained without regarding the algebraic signs of the measurements. Therefore, they have extensive applications in different fields, such as engineering, finance, insurance and so on. Folded exponential power (FEP) distribution is a newly proposed distribution which has modeling flexibility and easy usage [1]. In this study, we therefore consider different parametric methods for estimating the unknown parameters of FEP distribution. Maximum likelihood (ML), ordinary and weighted least squares (LS and WLS), Cramer von Mises (CVM) and maximum product of spacings (MPS) methods are used during the estimation process. The performances of the considered estimators are compared in a Monte-Carlo simulation study via bias and mean squared error (MSE) criteria. Results show that MPS method outperforms its rivals. Two real life applications taken from the literature are also considered.