|
Activity Number:
|
420
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 1, 2007 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
ENAR
|
| Abstract - #308464 |
|
Title:
|
Robust Estimation in the Presence of Extreme Censoring
|
|
Author(s):
|
Ruta Bajorunaite*+ and Vytaras Brazauskas
|
|
Companies:
|
Marquette University and University of Wisconsin-Milwaukee
|
|
Address:
|
PO Box 1881, Milwaukee, WI, 53201,
|
|
Keywords:
|
survival data ; censoring ; parametric models ; robust estimation
|
|
Abstract:
|
Parametric models are frequently used in modeling survival data. In many studies we encounter data where a certain proportion of extreme observations is censored. Maximum likelihood estimation is a standard tool used to fit parametric models. However, when the underlying model is mis-specified or contaminated the maximum likelihood parametric methods may be severely affected and lead to very poor results. We propose a robust parametric model fitting procedure for continuous failure time models. The procedure is based on trimmed L-statistics and it can achieve various degrees of robustness, which can be easily specified by the user as trimming proportions. Unlike most M-estimators, the newly developed method is straightforward to implement in practice as it (typically) does not require numerical solution of non-linear equations. The procedure is illustrated using real data example.
|
