多个指数分布参数基于截尾样本在序约束下的极大似然估计

Maximum Likelihood Estimation of Parameters of Some Exponential Distributions Based on Censored Samples under Order Restrictions

对给定k(k>2)个指数分布总体,抽取一组截尾样本1,2,i i iniX X X,讨论其均值θi,(i=1,2,3 k)在简单半序、简单树半序、伞型半序和简单环半序约束条件下的极大似然估计问题,根据PAVA算法给出了k个指数分布参数基于截尾样本在序约束下极大似然估计的计算方法。

Maximum likelihood estimation proplems on parameters of k exponential distributions based on censored samples under a simple half-order restriction,a simple tree order restriction,an umbrella order restriction and a simple ring order restriction were all discussed.With the PAVA algorithm,the iterative algorithm of maximum likelihood estimation of parameters of k exponential distributions under the order restrictions was given.