Probabilistic models of finite mixtures with components Generalized Extremal Value (GEV) distributions are widely applied in diverse areas such as finance and hydrology. In this work, the estimators of the parameters of the GEV mixture of two components are obtained via the EM algorithm. We also present numerical examples of the behavior of the estimates obtained through of simulation, well as an application to real data.
GEV; finite mixture; EM algorithm
