双定时混合截尾下双参数指数分布参数的Bayes估计

打开文本图片集

中图分类号：0213 文献标志码：A 文章编号：1673-5072（2025）05-0492-（

Bayes Estimation for Parameters of Two-parameter Exponential Distribution Under Double Type-I Hybrid Censoring Scheme

MU Yue-jia，LI Yun-fei

（School of Mathematics& Information，China West Normal University，Nanchong，Sichuan 637oo，China）

Abstract：Under the double type-I hybrid censoring test，the maximum likelihood estimations of the gate parameter （μ） and the scale parameter （θ） is firstly obtained for the two-parameter exponential distribution.Then the Bayes estimation and E-Bayes estimation of θ under three loss functions when μ is known are discussed.And the optimal values of the hyperparameters in the conjugate prior distribution are obtained by combining with the genetic algorithm，and the numerical simulation shows that the genetic algorithm can estimate the hyperparameters more accurately，thus improving the accuracy of the parameter Bayes estimation. Finally，the Bayes estimation of two parameters when both μ and θ are unknown is investigated，and numerical simulations show that the Bayes estimation accuracy of two parameters is higher than the maximum likelihood estimation accuracy of them.

Keywords：two-parameter exponential distribution； double type-I hybrid censoring；maximum likelihood estimation；Bayes estimation；E-Bayes estimation

双参数指数分布在多个领域都有独特的应用价值，尤其在需要描述和分析随机事件发生时间间隔的场景，如寿命数据分析、可靠性研究以及金融领域的保险损失分析等方面。（剩余6426字）