BARBOSA, D. G.; BARBOSA, Deyvison Guimarães.
Résumé:
The robust estimation method is an important tool to limit the influence of outliers
on both the response variable and the covariates of an accelerated life model.
In this work we used the Birnbaum-Saunders distribution, which has been extensively
applied in survival analysis studies. We presented the log-linear form of the accelerated
life model, we performed the robust estimation and approximated the variances of
the robust estimators by a function of variance and covariance of the sandwich type.
In addition, using different scenarios, we performed simulation studies to evaluate the
performance of the proposed robust estimator. Finally, we present two applications in
real data sets that illustrate the methodology presented.