基于Laplace分布下混合联合位置与尺度模型的参数估计

Parameters Estimation for Mixture of Joint Location and Scale Models Based on the Laplace Distribution

Laplace分布是分析厚尾数据的重要统计工具之一,本文基于Laplace分布提出了稳健的混合联合位置和尺度参数的回归模型,通过EM算法给出了该模型参数的极大似然估计,通过随机模拟试验验证了所提出方法的有效性.本文结合实际数据说明了该模型和方法具有实用性和可行性.

Laplace distribution is one of the most important statistical data analysis tools in the datasets with peaked and heavy-tailed errors. We propose a mixture joint location and scale models based on the Laplace distribution. EM algorithm is considered to estimate unknown parameters in the maximum likelihood estimate procedure. The efficiency of models and method are proved through simulation. Furthermore, a real example is presented to illustrate the proposed methodology.