摘要
Pólya分布在气候统计中常用来拟合雾、雷暴等.本文给出了Pólya分布总体在全样本场合下参数的矩估计和极大似然估计,并研究了估计的存在性,并通过大量的Monte Carlo模拟说明了估计的精度,认为在样本较大的情形下极大似然估计优于矩估计.最后通过具体的雾与雷暴等气候统计数据说明本文方法的可行性.
Polya distribution can be used to simulate frog. thunderstorm, ect. in the climatic statistics. In the present paper, we obtain the moment estimate and maximum likelihood estimate of Polya distribution under full sample size case, the existence of the estimators must be satisfied a certain condition because of the restriction on the values of parameters. Thereafter we study the existence of the estimate. Also in this paper, we show the accuracy of the estimate by a large amount of Monte-Carlo simulation and we find out that the maximum likelihood estimate is better than moment estimate under larger sample size case. Finally we illustrate the feasibility of this method by the detailed climatic statistical data, such as frog. thunderstorm, ect.
出处
《数理统计与管理》
CSSCI
北大核心
2008年第2期215-226,共12页
Journal of Applied Statistics and Management
基金
上海市教委基金(06MS009、CL200517)
上海市重点学科(T0401)
国家自然科学基金(10571057)
2007年度市教委重点课程(5Z1206)
2007年度统计学专业建设(5Z1501)
上海市科委科技项目(075105118)资助
关键词
Polya分布
矩估计
极大似然估计
气候统计
雾
雷暴
Polya distribution
moment estimate
maximum likelihood estimate
climatic statistics
frog
thunderstorm.
