截尾变量独立、不必同分布的条件下,最大似然估计的相合性及渐近正态性

CONSISTENCY AND ASYMPTOTIC NORMALITY OF MAXIMUM LIKELIHOOD ESTIMATORS FOR INDEPENDENT NOT IDENTICALLY DISTRIBUTED CENSORING SAMPLES

摘要: 本文提出条件（φ）,并证明了,在条件（φ）下,关于寿命分布的参数的最大似然估计具有强相合性及渐近正态性。验证了指数分布、Weibull分布、对数正态分布满足条件（φ）。这说明（φ）是比较广泛适用的。

Abstract: In this paper, Condition(φ) is proposed and under this condition we prove Maximum Likelihood Estimators for life distributed parameter is strong consistent and asymptotic normal. Then, we verify that Exponential distribution, Weibull distribution and Lognormal distribution are statified with Condition(φ) . It shows that Condition(φ) is widely applicable.