STATISTICAL METHODS FOR AIDS PREVALENT COHORT DATA

点击翻译按钮获取中文摘要

The objective of this grant is to develop statistical methods useful for prevalent cohort data in AIDS studies. The focus will be on two types of prevalent cohorts: (1) prevalent cohort with know time-origins, and (2) prevalent cohort with unknown time-origins. These two types of cohorts share one thing in common, namely, patients in both of the cohorts are recruited according to a certain cross-sectional sampling criterion. However, the data informations differ due to the known and unknown time- origins. The major topics will include non- and semi-parametric methods for one- sample and regression problems. In particular, statistical methods and theory for survival curve estimation, HIV-infection estimation, bootstrapping procedures and proportional hazard model will be developed.