首页 | 本学科首页   官方微博 | 高级检索  
     

聚类抽样情形下基于广义线性模型的风险比估计
引用本文:邓伟奇,张忠占. 聚类抽样情形下基于广义线性模型的风险比估计[J]. 应用数学, 2004, 0(Z2)
作者姓名:邓伟奇  张忠占
作者单位:北京工业大学应用数理学院 北京100022(邓伟奇),北京工业大学应用数理学院 北京100022(张忠占)
基金项目:北京市留学回国人员择优资助项目,国家自然科学基金资助项目(10371005)
摘    要:Liu等人在聚类抽样的情形下基于贝塔二项分布模型讨论了风险比的区间估计问题.本文以他们的研究为背景,建立了一类聚类抽样样本的广义线性模型,并采用拟似然方法导出相应的广义估计方程,从而获得风险比的直接估计.这种方法不需要假定分布的具体形式,因而比Liu等人的方法具有更好的适用性.最后在一定正则条件下证明了估计的相合性和渐近正态性质.

关 键 词:拟似然估计  广义估计方程  相合性  渐近正态性

Estimation of the Risk Ratio in Cluster Sampling Based on Generalized Linear Model
DENG Wei qi,ZHANG Zhong zhan. Estimation of the Risk Ratio in Cluster Sampling Based on Generalized Linear Model[J]. Mathematica Applicata, 2004, 0(Z2)
Authors:DENG Wei qi  ZHANG Zhong zhan
Abstract:Liu,Mayer and Eckhardt discussed interval estimtion of the RR in cluster sampling on the basis of beta binomial model.For the similar background to their research,this paper focuses its aim on the direct estimation of RR through a generalized linear model for clustered data where the method of quasi likelihood is employed to reach a generalized estimating equation for that purpose.Since no special assumption on distribution forms is needed,the way we adopt is more applicable than that of those authors mentioned above.Under some regular conditions,we get the consistency and asymptotic normality of the estimator.
Keywords:Asymptitic normality  Consistency  Generalized estimation equation  Quasi-likelihood estimation
本文献已被 CNKI 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号