Volume 13, January 2009
|Page(s)||87 - 114|
|Published online||26 March 2009|
Estimation of the hazard function in a semiparametric model with covariate measurement error
UMR AgroParisTech/INRA MIA 518, Paris, France.
2 URGV UMR INRA 1165/CNRS 8114/UEVE, Évry, France.
3 Université Paris Descartes, Laboratoire MAP5, Paris; email@example.com
Revised: 30 January 2008
We consider a failure hazard function, conditional on a time-independent covariate Z, given by . The baseline hazard function and the relative risk both belong to parametric families with . The covariate Z has an unknown density and is measured with an error through an additive error model U = Z + ε where ε is a random variable, independent from Z, with known density . We observe a n-sample (Xi, Di, Ui), i = 1, ..., n, where Xi is the minimum between the failure time and the censoring time, and Di is the censoring indicator. Using least square criterion and deconvolution methods, we propose a consistent estimator of θ0 using the observations n-sample (Xi, Di, Ui), i = 1, ..., n. We give an upper bound for its risk which depends on the smoothness properties of and as a function of z, and we derive sufficient conditions for the -consistency. We give detailed examples considering various type of relative risks and various types of error density . In particular, in the Cox model and in the excess risk model, the estimator of θ0 is -consistent and asymptotically Gaussian regardless of the form of .
Mathematics Subject Classification: 62G05 / 62F12 / 62G99 / 62J02
Key words: Semiparametric estimation / errors-in-variables model / measurement error / nonparametric estimation / excess risk model / Cox model / censoring / survival analysis / density deconvolution / least square criterion
© EDP Sciences, SMAI, 2009
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.