基于删失试验的贝叶斯共享异质分段指数危险模型

Bayesian Different Sub-index Risk Sharing Frailty Model Based on Censored Test

针对在删失试验的生存分析中,为了估计不同协变量在组内的相互影响以及一些无法观测的协变量产生的异质影响,在基准危险函数为分段指数的情形下,给出贝叶斯共享异质模型,利用Gibbs抽样得出参数的后验分布。然后对模型给出一个实证分析.模型采用乘法异质模型,利用WinBUGS软件得出后验参数相关统计量,说明此模型的有效性和可靠性.

In view of the censored test of survival analysis, in order to estimate different variables in the group＇s interaction and a number of variables can not be observing, the Association for the heterogeneity impact of the baseline risk of sub - index for the function of the circumstances, given Bayesian sharing heterogeneous model, using the parameters that Gibbs sampling inspection after the distribution, then the model gives an empirical analysis. Models use multiplication heterogeneous model, using WinBUGS software, access to posterior parameters relevant statistics, itwas to prove that model of effectiveness and reliability.