|
|||||
|
|
| 左截断右删失数据下指数分布参数多变点的贝叶斯估计 | |
| 引用本文： | 何朝兵[],、刘跃军[],、刘华文[].左截断右删失数据下指数分布参数多变点的贝叶斯估计[J].西南师范大学学报(自然科学版),2015,40(1). |
| 作者姓名： | 何朝兵[] 、刘跃军[] 、刘华文[] |
| 作者单位： | 1. 安阳师范学院数学与统计学院,河南安阳,455000 2. 安阳师范学院软件学院,河南安阳,455000 3. 山东大学数学学院,济南,250100 |
| 基金项目： | 国家自然科学基金项目(61174099)；河南省教育厅科学技术研究重点项目(12A520001)． |
| 摘 要： | 主要利用MCMC方法研究了左截断右删失数据下指数分布多变点模型的参数估计问题.通过筛选法和逆变换法得到了指数分布的完全数据,在获得各参数的满条件分布后,利用MCMC方法得到了Gibbs样本,把Gibbs样本的均值作为各参数的估计.随机模拟的结果表明各参数估计的精度都较高. |
| 关 键 词： | 完全数据似然函数 满条件分布 MCMC方法 Gibbs抽样 Metropolis-Hastings算法 |
On Bayesian Estimation of Parameter of Exponential Distribution with Multiple Change Points for Randomly Truncated and Censored Data |
|
| HE Chao-bing,LIU Yue-jun,LIU Hua-wen.On Bayesian Estimation of Parameter of Exponential Distribution with Multiple Change Points for Randomly Truncated and Censored Data[J].Journal of Southwest China Normal University(Natural Science),2015,40(1). | |
| Authors: | HE Chao-bing[] LIU Yue-jun[] LIU Hua-wen[] |
| Abstract: | Parameter estimation in exponential distribution multiple change points model for randomly trun-cated and censored data has been studied in this paper .The complete data of exponential distribution has been obtained by screening method and inverse transformation method .After the full conditional distribu-tions of all parameters have been obtained ,Gibbs samples have been obtained in MCMC method ,and the means of Gibbs samples has been taken as Bayesian estimations of the parameters .Random simulation re-sults show that Bayesian estimations of the parameters are fairly accurate . |
| Keywords: | complete-data likelihood function full conditional distribution MCMC method Gibbs sam-pling Metropolis-Hastings algorithm |
| 本文献已被 CNKI 万方数据 等数据库收录！ | |
| 点击此处可从《西南师范大学学报(自然科学版)》浏览原始摘要信息 | |
| 点击此处可从《西南师范大学学报(自然科学版)》下载免费的PDF全文 | |