Accelerated Failure Time Model Analysis For Current Status Data

In many scientific fields,such as medicine,biology,demographic,we often can not observed the accurate failure time of the research object.What can be measured is whether the failure happened before or after the observation time.This kind of data is called Current status data.At present,some achievements have been made in the research of Current status data.In this paper,we study the related estimation question about the accelerated failure time model of Current status data,and introduce the method of unbiased transformation to estimate the regression parameters in the accelerated failure time model.The estimation of distribution function of the random error term is obtained by the method of isotonic regression.The consistency of the estimation of distribution function of the random error term is proved,The feasibility of the proposed method is verified by the simulation results.