|
Abstract
|
The weighted xgamma distribution, a new weighted two-parameter lifespan distribution, is introduced in this
study. Theoretical characteristics of this model are deduced and thoroughly examined, including quantile function, extreme
value, moments, moment generating function, cumulative entropy, and residual cumulative. Some classical estimation
methods such as the the maximum likelihood, weighted least square, Anderson Darling and Cramer-von-Mises are
considered. A simulation experiments are performed to compare the estimation methods. Four real-life data sets is finally
examined to demonstrate the viability of this model. Four key risk indicators are defined and analyzed under the maximum
likelihood method. A risk analysis for the exceedances of flood peaks is presented.
|