融合面板计数数据下相依区间删失数据的参数估计  

Parameter Estimation of Dependent Interval-Censored Data and Panel Count Data

作　　者：王淑影 汪童 黄鹤 WANG Shuying;WANG Tong;HUANG He(School of Mathematics and Statistics,Changchun University of Technology,Changchun 130000)

机构地区：[1]长春工业大学数学与统计学院,长春130000

出　　处：《系统科学与数学》2025年第2期613-629,共17页Journal of Systems Science and Mathematical Sciences

基　　金：吉林省自然科学基金(20230101371JC)资助课题。

摘　　要：区间删失数据和面板计数数据是生存分析中常见的两类不完整数据.文章考虑两类数据的联合建模分析,同时纳入了带有信息的观测过程.引入两个脆弱项来刻画失效时间、计数过程和观测过程的相关性,并对三者进行联合建模研究,利用两步估计法实现所建模型的参数估计.然后进行数值模拟,模拟结果表明所提出的方法表现良好.最后将所提出的方法应用于同种异体心脏移植血管病变研究的真实数据.Interval censored data and panel count data are two types of incomplete data that often appear in survival analysis.In this paper,we consider the joint modeling analysis of the two types of data,also the dependent observation process is included.Then we introduce two frailties to characterize the correlation among the failure time,counting process and observation process.This study conducts a joint modeling study among the three components.The two-step estimation procedure is proposed to realize the parameter estimation of the established model.Then,numerical simulation is carried out,and the simulation results show that the proposed method performs well.Finally,the proposed method is applied to the real data of cardiac allograft vasculopathy research.

关 键 词：区间删失数据 面板计数数据 联合模型 极大似然估计 两步估计法 

分 类 号：O212.1[理学—概率论与数理统计]