Semiparametric additive accelerated life models

In this paper we investigate the asymptotic properties of estimators obtain ed for the scmiparametric additive accelerated life model proposed by Bagdo navicius & Nikulin (1995). This model generalizes the well known additive h azards model of survival analysis and is close to the general transformatio n model (sce Dabrowska & Doksum, 1988). Asymptotic properties of the estima tor of the regression parameter and the estimator of the reliability functi on are given in the case of right censoring for discretized data and a nume rical example illustrates these results.