所属单位：数学与统计学院
教研室：统计教研室
发表刊物：Lifetime Data Analysis
项目来源：国家自然科学基金项目
关键字：Competing risks , Copula models, Current status data, Nonparametric
摘要：Nonparametric estimation of a survival function is often the first task in the analysis of failure time data and this paper discusses this problem when one observes only current status data (McKeown and Jewell, 2010; Sun, 2006). In this situation, each subject is observed only once and the failure time of interest is either left- or right-censored. If the failure time and the observation time can be assumed to be independent, the maximum likelihood estimate of the survival function can be easily derived. However, the independent assumption may not hold in practice and with dependent observation time, it becomes a competing risk problem although not a standard one with current status data.For the problem, we employ a copula model, a natural and commonly used approach for competing risks, and show that if the copula function is known, the survival function of interest can be consistently and uniquely estimated. Two simple estimates are proposed and an simulation study is performed to assess their performance.
合写作者：Sun,Jianguo,孙六全,周杰,王德辉
第一作者：王纯杰
论文类型：期刊论文
卷号：18
期号：4
页面范围：1
ISSN号：1380-7870
是否译文：否
发表时间：2012-10-01