Document Type
Lecture
Publication Date
2-13-2004
Abstract
In Koziol-Green (KG) model, the hazard of censoring time is assumed to be proportional to that of survival time. One of the approaches on estimating the survival function in the literature is the so-called Abdushukurov-Cheng-Lin (ACL) estimation. In this presentation, the generalized maximum likelihood method (GMLE) will be introduced and the small and large sample properties of GMLE comparing with ACL and Kaplan-Meier (KM) estimators will be explored. Furthermore, the asymptotic efficiency of GMLE and ACL estimators and the generalization of KG model will be discussed. The main results from our recent research can be summarized as follow. 1) In both small and large samples, GMLE and ACL estimators always outperform KM estimator; 2) GMLE and ACL estimators are asymptotically equivalent; 3) They are asymptotically efficient in the sense of being the least dispersed regular estimators.
Relational Format
presentation
Recommended Citation
Zhang, Haimeng, "Estimating Survival Functions In Koziol-Green Models" (2004). Probability & Statistics Seminar. 67.
https://egrove.olemiss.edu/math_statistics/67
Accessibility Status
Searchable text
