题　目：Nonparametric Analysis of Doubly Truncated and Interval Censored Data 主讲人：沈葆圣 教授(东海大学统计系) 时　间：99年3月26日(星期五)上午10:40-11:30 (上午10:20-10:40茶会於交大统计所429室举行) 地　点：交大综合一馆427室 Abstract Doubly truncated data appear in a number of applications, including astronomy and survival analysis. For double-truncated data, the lifetime T is observable only when U ≦T≦V, where U and V are the left-truncated and right-truncated time, respectively. In this talk, we will review the existing numerical algorithms and technical results (Efron and Petrosian (1999) and Shen (2008)) on the analysis of doubly truncated data. In some situation, the lifetime T also suffers interval censoring. Under the assumption of independence between T and (U, V), using EM algorithm of Turnbull (1976), we propose a nonparametric estimate of the distribution function of T. Simulation results indicate that the proposed estimator performs adequately when truncation is not severe. However, difficulties arise in trying to derive the asymptotic properties of the nonparametric maximum likelihood estimate (NPMLE). The consistency properties of the NPMLE will be discussed. Key Words: double truncation; interval censoring; nonparametric maximum likelihood estimation. -- --

※ 发信站: 批踢踢实业坊(ptt.cc) ◆ From: 140.113.180.30