Parameter Estimation For The Accelerated Failure Time Model About Multivariate Two-stage Current-status Data

Recently,as an alternative and supplement to the proportional hazard model,the accelerated failure time model development very rapidly, relevant research results are more and more,but the research about multivariate accelerated failure time model(AFT) is very little, especially the current-status data. The importance and practicality of the AFT model is beyond doubt, which is the theoretical significance of this paper.In practice, our survival data are often very complex, the current-status two-stage data is one of the more important one, have very important applications in medical statistics and bio statistics.This paper discusses the method and some related simulation and test parameters of the accelerated failure time model of two phase estimation.We combine the form of the general accelerated failure time model and the special structure of the two stage of the current data to introduce a multivariate accelerated failure time model for the analysis of the two-stage data.In view of some difficult problems faced by maximum likelihood estimation, a new algorithm is proposed, which is based on the combination of MC and NR algorithm.Then, we have carried on some numerical simulation to the model, and compared the simulation results of the two stage data from the model and the two kinds of other methods, the results show that our method is more advantageous.Finally, the robustness of the model is analyzed and a simple model test is carried out to prove the accuracy of our model, and based on this, the model is extended to a more general situation.