期刊导航

论文摘要

三参数威布尔分布贝叶斯估计的混合Gibbs算法

Mixed Gibbs algorithm of Bayesian estimation for three parameter Weibull distribution

作者:魏艳华(天水师范学院数学与统计学院);王丙参(天水师范学院数学与统计学院);孙永辉(河海大学能源与电气学院)

Author:WEI Yan-Hua(School of Mathematics and Statistics, Tianshui Normal University);WANG Bing-Can(School of Mathematics and Statistics, Tianshui Normal University);SUN Yong-Hui(College of Energy and Electrical Engineering, Hehai University)

收稿日期:2013-06-25          年卷(期)页码:2015,52(2):233-240

期刊名称:四川大学学报: 自然科学版

Journal Name:Journal of Sichuan University (Natural Science Edition)

关键字:完全样本; 定数截尾样本; 三参数威布尔分布; 贝叶斯估计; 混合Gibbs算法

Key words:Complete sample; Type II censored sample; Three parameter Weibull distribution; Bayesian estimation; Mixed Gibbs algorithm

基金项目:国家自然科学基金(61104045); 天水师范学院中青年教师科研资助项目(TSA1404)

中文摘要

利用混合Gibbs算法(Gibbs抽样与Metropolis算法的混合)给出了完全样本和定数截尾样本两种情形下三参数威布尔分布的贝叶斯估计. 作为应用举出了三个实例, 通过Monte Carlo模拟求得参数的贝叶斯估计及可信区间, 给出混合Gibbs抽样过程中参数的轨迹图、直方图及自相关系数图, 并将模拟结果与相关文献进行了比较

英文摘要

we propose a Bayesian estimation for three parameter Weibull distribution by using a mixed Gibbs algorithm (Gibbs sampling with Metropolis algorithm mixed) both in the case of complete sample and type II censored sample. Then, as some applications, we take three examples, in which we obtain Bayesian estimation and the confidence interval for the parameters by the Monte Carlo simulation,gives locus respectively, histograms and autocorrelation coefficient figure of the parameters in the mixed Gibbs sampling process. Then the simulation results are compared with some known results.

关闭

Copyright © 2020四川大学期刊社 版权所有.

地址:成都市一环路南一段24号

邮编:610065