基于删失数据的Log-Gompertz回归模型的统计推断（英文）

Statistical Inference of Log-Gompertz Regression Model with Censored Data

Gompertz模型在人类学研究中被广泛使用.讨论了基于对数不换的Gompertz模型,即log-Gompertz模型.研究了参数的极大似然估计和渐近置信区间,并提供了利用Bootstrap方法寻找参数的置信区间.在不同参数、不同样本容量以及不同删失计划的情形下,给出相关数值模型结果用以评价所建立的回归模型.

The Gompertz distribution is widely applied in modeling human mortality. In this article, the regression model is discussed, which is based on the logarithm of the Gompertz distribution, namely, log-Gompertz distribution. The method of maximum likelihood for estimating the model parameters is studied. The bootstrap method is also applied to constructing the confidence interval of parameters. For different parameter settings, sample sizes, censoring percentages and various simulation studies are performed to assess the regression models.