(a,b,0)类分布族及其最大后验密度可信区间

(a,b,0)Class Distributions and Their Confidence Interval of Highest Posterior Density

为了在模拟索赔次数时更好的运用(a,b,0)类分布族,系统研究了(a,b,0)类分布族的优良特性及相互关系,利用贝叶斯统计分析方法,在取先验分布为共轭分布的情形下,给出了最大后验密度可信区间,即最短可信区间,并通过实例与经典区间估计进行了比较.

This paper systematically discusses the properties and relationships among families to use them in the stochastic simulation. When prior distributions are conjugate distributions, it gives their confidence interval of the highest posterior density by using the method of Bayes statistical analysis, and compares the classical interval estimation by examples.