改进自适应渐进 II 型删失下 Chen 分布的贝叶斯分析  

Bayesian Analysis of Chen Distributionunder Improved Adaptive ProgressiveType-II Censoring

作　　者：张莉 

机构地区：[1]西北师范大学数学与统计学院,甘肃 兰州

出　　处：《理论数学》2024年第1期400-415,共16页Pure Mathematics

摘　　要：本文基于改进的自适应渐进 II 型删失,对 Chen 分布进行了贝叶斯分析。首先利用 EM 算法得 到了参数的极大似然估计。针对共轭和非共轭的四种信息先验,运用大方差和遗传算法确定了先 验的超参数。进而根据 Metropolis-Hastings 算法实现了后验分布样本的抽取。最后,通过真实 数据集对不同先验下的贝叶斯估计性能进行比较并得出了相应的结论。This paper presents a Bayesian analysis of the Chen distribution based on an improvedadaptive progressive Type-II censoring. Initially, the EM algorithm is used toobtain the maximum likelihood estimation of the parameters. Four types of informativepriors, including conjugate and non-conjugate, are then determined using largevariance and genetic algorithms. Subsequently, the Metropolis-Hastings algorithm isemployed to extract samples from the posterior distribution. Finally, a comparisonof the Bayesian estimation performance under different priors is conducted using realdatasets, leading to corresponding conclusions.

关 键 词：改进的自适应渐进 II 型删失 Chen 分布 EM 算法 共轭和非共轭的信息先验 大方差和遗传 算法 Metropolis-Hastings 算法 贝叶斯估计 

分 类 号：TP3[自动化与计算机技术—计算机科学与技术]