带有信息的区间删失数据下AFT模型的参数估计  

Parameter Estimation of AFT Model with Informatively Interval-Censored Data

作　　者：王淑影 刘舒也 张亚男 程云飞 WANG Shu-ying;LIU Shu-ye;ZHANG Ya-nan;CHENG Yun-fei(School of Mathematics and Statistics,Changchun University of Technology,Changchun 130012,China)

机构地区：[1]长春工业大学数学与统计学院,吉林长春130012

出　　处：《数理统计与管理》2024年第1期58-65,共8页Journal of Applied Statistics and Management

基　　金：国家自然科学基金青年项目(11901054);国家自然科学基金数学天元基金资助项目(11926340,11926341);中国博士后科学基金面上项目(2021M700536)。

摘　　要：在生存分析研究中,多数文章假定感兴趣的失效时间和删失时间是独立的,但这一假设在实际情况中未必合理。如果忽略失效时间与删失时间的相依性,可能会导致错误的结论。所以本文考虑在带有信息的K型区间删失数据下,采用基于两步估计的极大似然估计方法对误差项服从标准正态分布的加速失效时间模型(accelerated failure time model,AFT)进行参数估计。同时还进行了数值模拟以验证提出方法的有效性。最后,应用所提出的方法分析艾滋病的临床试验数据。For the research of survival analysis,most of existing approaches assume that the censored time is independent of the failure time of interest and it is clear that this may not be true in practice.If we ignore the dependence between the failure time of interest and the censored time,which may cause wrong conclusions.In this article,we consider a general type of censored data,the informatively case K interval-censored data.For the problem,a maximum likelihood estimation approach based on a two-step procedure is proposed to estimate the parameters of accelerated failure time(AFT) model whose error item follows the standard normal distribution.A simulation study is conducted to assess the effectiveness of the proposed method and indicates that it works well.Finally,the method is applied to a set of real data from AIDS clinical trail.

关 键 词：加速失效时间模型 带有信息的K型区间删失数据 极大似然估计法 

分 类 号：O212[理学—概率论与数理统计]