摘要
本文研究了Pareto分布中形状参数的估计问题.采用极大似然估计及其渐近正态性、轮廓似然估计和矩估计的方法对形状参数估计,通过生成Pareto分布随机数的方法进行数值模拟,比较渐近正态性和轮廓似然估计得到的区间长度以及极大似然估计和矩估计得到的估计值,给出各估计的优缺点.
The methods including maximum likelihood estimation,asymptotic normality,profile likelihood estimation,and moment estimation were applied to study shape parameter estimation in Pareto distribution in this paper. The digital simulation was performed by generating Pareto-distributed random numbers. The interval lengths obtained from the asymptotic normality and profile likelihood estimation were compared,and the estimation values from maximum likelihood estimation and moment estimation were also compared. Finally,the merits and demerits of various estimations were given.
引文
[1]刘荣玄,邬四英,张萍.关于三参数Pareto分布的参数估计问题[J].井冈山大学学报,2015,36(4):7~12.
[2]王娟,徐付霞.Pareto分布中尺度参数的区间估计[J].哈尔滨商业大学学报,2015,31(5):629~632.
[3]史道济.实用极值统计[M].天津:天津科技技术出版社,2005.
[4]茆诗松,王静龙.高等数理统计[M].北京:高等教育出版社,2004.