混合双参数广义Pareto分布的参数估计

Estimation of Parameters of Mixed Generalized Pareto Distribution

Pareto分布族因其厚尾特点,在金融分析、寿命分析中都是非常重要的统计模型.但是对于混合双参广义Pareto分布,在模型参数估计时,传统的矩法估计和极大似然估计在理论上可以实现,实践时比较困难.本文应用EM算法之ECM算法,研究了混合广义Pareto分布在完全数据场合下的参数估计问题,并模拟说明EM算法来估计混合广义Pareto分布是一种容易实现又非常有效的方法.

Generalized Pareto distribution plays an important role in financial and life time data analysis for its fat-tailed. However we may encounter enormous difficulties when we try to estimate the model by traditional methods such as maximize likelihood （ML） estimation and moment estimation, though they make sense theoretically. In this paper, we employ ECM algorithm of an extension EM algorithm to obtain the parameter estimation for the mixed Generalized Pareto model with complete sample. Simulations are also made to demonstrate the efficiency of this estimation.