删失数据情形置信下限的模型选择方法

A MODEL SELECTION METHOD FOR LOWER CONFIDENCE LIMIT IN CENSORED CASE

本文就一般总体未知参数的置信下限的设置问题提出了一种模型选择方法，其中数据是定时截尾的．所提出的方法可推广到一般情形．

In this paper, a model selection method is established for setting the lower confidence limit ofthe unknown parameter of a population. The data are censored at fixed times. Our method can be appliedto general case.