泊松分布参数的稳健估计

Robust Estimation of Parameter in Poisson Distribution

本文主要以非对称分布中泊松分布为研究对象,探讨了其参数的稳健估计方法作者以截断似然估计为基础,结合Cizek的工作,提出了适用于泊松分布参数的一种稳健估计方法.该方法避免了事先选取截断比例的麻烦,通过数据自身的信息给出在平均似然最大准则下的最优截断比例.在文中的模拟部分,分别就未受污染和受污染的泊松分布数据进行了模拟,得到了不错的效果.

This paper, the asymmetrical distribution of the Poisson distribution as an objective, discussed the estimationmethod of robust parameter. Author truncated likelihood estimation, combining Cizek＇s work, proposed a robust estimationmethod applying to Poisson distribution parameters. It avoided the hassle of pre-selected cutoff ratio, and gave theirinformation through the data at an average maximum likelihood ratio criterion optimal truncation. In the analog part of thetext, uncontaminated and contaminated Poisson distribution data were respectively simulated to get good results.