基于MCMC稳态模拟的Weibull共享异质性模型及其可靠性应用
1.南京理工大学经济管理学院,南京 210094
2.湖南大学统计学院,长沙 410079
下一篇 上一篇
摘要
针对传统假设中个体寿命独立同分布的不足,构建了贝叶斯Weibull共享异质性模型,提出了对寿命服从Weibull分布的产品,运用基于Gibbs抽样的马尔可夫链蒙特卡罗(Markov chain Monte Carlo,MCMC)方法动态模拟出参数后验分布的马尔可夫链,在异质性因子的先验分布为Gamma分布时,给出随机截尾条件下,参数在Weibull共享异质性模型中的贝叶斯估计,提高了计算的精度。借助数据仿真说明了利用WinBUGS(Bayesian inference using Gibbs sampling)软件包进行建模分析的过程,证明了该模型在可靠性应用中的直观性与有效性。
参考文献
[ 1 ] 彭非,王伟.生存分析[M].北京:中国人民大学出版社,2004
[ 2 ] VaupelJM,Manton K G,Stalland E.The impact of heterogeneity in individual frailty on the dynamics of mortality[J].Demography,1979,(16):439~454 链接1
[ 3 ] IbrahimJG,Chen Minghui,Sinha D.Bayesian Survival Analysis[M].New York Berlin Heidelberg,2001.13~17
[ 4 ] Clayton DG.A model for association in bivariate life tables and its application in epidemiological studies of familial tendency in chronic disease incidence[J].Biometrika,1978,(65):141~151 链接1
[ 5 ] Philip H.Analysis of Multivariate Survival Data[M].New York Berlin Heidelberg,2001
[ 6 ] Lee S,Klein JP.Bivariate models with a random environmental factor[J].Indian Journal of Productivity,Reliability and Quality Control,1988,(13):1~18
[ 7 ] Duchateau L,Jansen P,Lindsey P,et al.The shared frailty model and the power for heterogeneity tests in multicenter trials[J].Computational Statistic&Data Analysis,2002,(40):603~620 链接1
[ 8 ] dos Santos DM,Davies R B,Francis B.Nonparametric hazard versus nonparametric frailty distribution in modeling recurrence of breast cancer[J].Jourmal of Statistical Planning and Inference,1995,(47):111~127 链接1
[ 9 ] Yashin A1,lachine I A.Dependent hazards in multivariate survival problems[J].Jourmal of Multivariate Analysis,1999,(71):241~261 链接1
[10] Scollnik PM.Actuarial modeling with MCMC and BUGS[J].North American Actuarial Jourmal,2001,5(2):%~125 链接1
[11] Spiegelhaler D.Thomas A,Best N.BUGS 0.5:Bayesian Inference Using Gibbs Sampling Manual(Version i)[M].Cambridge,U K.:MRC Biostatistics Unit,1996
[12] LawlessJF.寿命数据中的统计模型与方法[M].茄诗松,濮晓龙,刘忠译,北京:中国统计出版社,1998