In this paper we combine empirical likelihood and estimating functions for censored data to obtain robust confidence regions for the parameters and more generally for functions of the parameters of distributions used in lifetime data analysis. The proposed method works with type I or type II or randomly censored data. It is illustrated referring to inference for log-location-scale models. In particular, we focus on the log-normal and the Weibull models and we consider the problem of constructing robust confidence regions (or intervals) for the parameters of the model, as well as for quantiles and values of the survival function. The usefulness of the method is demonstrated through a Monte Carlo study and by examples on two lifetime data sets.
Robust confidence intervals for log-location-scale models with right censored data.
2003
Abstract
In this paper we combine empirical likelihood and estimating functions for censored data to obtain robust confidence regions for the parameters and more generally for functions of the parameters of distributions used in lifetime data analysis. The proposed method works with type I or type II or randomly censored data. It is illustrated referring to inference for log-location-scale models. In particular, we focus on the log-normal and the Weibull models and we consider the problem of constructing robust confidence regions (or intervals) for the parameters of the model, as well as for quantiles and values of the survival function. The usefulness of the method is demonstrated through a Monte Carlo study and by examples on two lifetime data sets.File | Dimensione | Formato | |
---|---|---|---|
WP_2003_15.pdf
accesso aperto
Licenza:
Accesso gratuito
Dimensione
45 MB
Formato
Adobe PDF
|
45 MB | Adobe PDF | Visualizza/Apri |
Pubblicazioni consigliate
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.