首页> 外文期刊>International Journal of Statistics and Applications >A Comparison of Three Different Procedures for Estimating Variance Components
【24h】

A Comparison of Three Different Procedures for Estimating Variance Components

机译:三种估计方差分量的方法的比较

获取原文
           

摘要

As a consequence of various theoretical developments, and of improvements in computing strategies, restricted maximum likelihood (REML) estimation has become a viable procedure for estimating the variance components in mixed linear models. In this article, the procedure of Xu and Atchley (1996) has been extended in such a way that can be applied to any general linear model. Further, alternative estimator based on empirical Bayes approach has been derived for estimating the random- effects variance components in the light of REML function. Comparison between the proposed estimator and the estimators provided by Xu and Atchley (1996) and Moghtased-Azar et al. (2014) has been computed under unbalanced nested-factorial model with two fixed crossed factorial and one nested random factor. Finally, all the estimators in the vignette are activated by an illustrative example.
机译:由于各种理论的发展和计算策略的改进,有限最大似然(REML)估计已成为估计混合线性模型中方差分量的可行方法。在本文中,对Xu和Atchley(1996)的过程进行了扩展,使其可以应用于任何通用线性模型。此外,已经推导了基于经验贝叶斯方法的替代估计器,用于根据REML函数估计随机效应方差分量。建议的估计量与Xu和Atchley(1996)和Moghtased-Azar等人提供的估计量之间的比较。 (2014)是在具有两个固定交叉因子和一个嵌套随机因子的不平衡嵌套因子模型下计算的。最后,通过说明性示例激活了小插图中的所有估计器。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号