其未知参数估计通常用极大似然估计(M ax im um L ike li-hood E stim ates
M LE)
但有时极大似然估计不一定存在.S ilvapu lle and Burridge只给出了该模型极大似然存在的一个充分必要条件
但未给出严格的证明.本文给出其严格的证明
并且给出仅有区间截尾的情形下参数极大似然估计存在的一个更简单的条件.
Abstract
Linear regression models with censored data are frequently used in industrial statistics.In these models
maximum likelihood estimates (MLE) are employed to estimate unknown parameters though the MLE may not exist sometimes.Silvapulle and Burridge presented the necessary and sufficient conditions for identifying the existence MLE;they failed to offer strict reasoning.In this paper
the authors have presented a strict reasoning
and a condition for identifying the existence of MLE with interval censored.