Computation of Nonparametric Survival Functions in the Presence of Interval Censoring

Date
2008-11-18
Authors
Taylor, SM
Supervisor
Item type
Conference Contribution
Degree name
Journal Title
Journal ISSN
Volume Title
Publisher
NZMASP
Abstract

This talk will cover my research in developing an efficient and robust algorithm for solving a particular problem in survival analysis. The problem is to find the nonparametric maximum likelihood (NPMLE) survival function from time-to-event data where some or all of the observation times are censored by arbitrary intervals in the positive real line. The presence of this censoring makes the problem difficult; since no closed form solution is possible, iterative optimisation techniques must be used. We have called our new algorithm the “Hierarchical Constrained Newton Method” (HCNM) because it is based on the existing CNM algorithm, enhancing it with a divide and conquer approach.

Description
Keywords
Source
New Zealand Mathematics and Statistics Postgraduate Conference 2008, Whitianga, New Zealand
DOI
Rights statement
Auckland University of Technology (AUT) encourages public access to AUT information and supports the legal use of copyright material in accordance with the Copyright Act 1994 (the Act) and the Privacy Act 1993. Unless otherwise stated, copyright material contained on this site may be in the intellectual property of AUT, a member of staff or third parties. Any commercial exploitation of this material is expressly prohibited without the written permission of the owner.