Constrained Kaplan–Meier curve and empirical likelihood
Advanced Review
Published Online: May 03 2017
DOI: 10.1002/wics.1400
Can't access this content? Tell your librarian.
The Kaplan–Meier curve and the namesake Kaplan–Meier 1958 paper is the most cited statistical paper in the history. We study the error analysis of the popular Kaplan–Meier curve. We describe in detail and recommend a method based on the calculation of constrained Kaplan–Meier curve and the recently developed Empirical Likelihood theory. Examples are given illustrating how such error analysis are carried out, using the software R and package KMC. WIREs Comput Stat 2017, 9:e1400. doi: 10.1002/wics.1400 This article is categorized under: Statistical and Graphical Methods of Data Analysis > EM Algorithm Algorithms and Computational Methods > Maximum Likelihood Methods Algorithms and Computational Methods > Numerical Methods
A constrained Kaplan–Meier curve: two linear constraints at t = 500 and t = 900. A push up at t = 500 and pull down at t = 900.
[
Normal View
|
Magnified View
]