最大熵先验下的正态分布总体的Bayes分析  

Bayesian Analysis of Normal Distribution Under Maximum Entropy Priori

作　　者：王燕飞[1] 

机构地区：[1]吉林化工学院理学院,吉林吉林132022

出　　处：《数学的实践与认识》2017年第20期183-186,共4页Mathematics in Practice and Theory

摘　　要：在正态分布总体方差已知时,满足总体均值为正,且其期望为某常数的约束条件下,应用最大熵原理获取先验分布,并求得后验分布.利用该后验分布得出总体期望的Bayes估计及可信区间.解决了经典统计中难以解决的问题.最后用算例说明其应用.In this paper, when the total variance of normal distribution is known and oval mean is positive, the prior distribution is gained under the maximum entropy principle. And then the posterior distribution is obtained. Bayesian estimation and confidence interval of the overall expectation is obtained by using the posterior distribution. This solves the problem which is classic statistics hard to work. Finally, a numerical example is used to illustrate its application.

关 键 词：最大熵 正态分布 贝叶斯估计 贝叶斯可信区间 

分 类 号：O212.8[理学—概率论与数理统计]