程靖, 岳荣先. Panel Data模型基于Within估计的最优设计[J]. 应用概率统计, 2018, 34(1): 95-100. DOI: 10.3969/j.issn.1001-4268.2018.01.008
引用本文: 程靖, 岳荣先. Panel Data模型基于Within估计的最优设计[J]. 应用概率统计, 2018, 34(1): 95-100. DOI: 10.3969/j.issn.1001-4268.2018.01.008
CHENG Jing, YUE RongXian. Optimal Designs of Panel Data Model Based on Within Estimator[J]. Chinese Journal of Applied Probability and Statistics, 2018, 34(1): 95-100. DOI: 10.3969/j.issn.1001-4268.2018.01.008
Citation: CHENG Jing, YUE RongXian. Optimal Designs of Panel Data Model Based on Within Estimator[J]. Chinese Journal of Applied Probability and Statistics, 2018, 34(1): 95-100. DOI: 10.3969/j.issn.1001-4268.2018.01.008

Panel Data模型基于Within估计的最优设计

Optimal Designs of Panel Data Model Based on Within Estimator

  • 摘要: 本文研究二元Panel Data模型基于Within估计在恒等设计类中的D-, A-, I-和E-\,最优设计. 文章首先证明了二元Panel Data模型基于Within估计在恒等设计类中的几类最优设计可在设计域的顶点处获得;研究发现模型的D-, A-和E-\,最优设计均为设计域顶点上的等权重设计,I-\,最优设计为对称设计; 结果还表明等权重设计具有较高的I-效.

     

    Abstract: In this paper, optimal identical designs of bivariatepanel data model based on within estimator are discussed. It is proved that D-, A-, E- and I-optimal design of bivariate panel data model can be obtained on the vertexes of the design region. Equal-weight design is proved to be D-,A-, E-optimal of bivariate panel data model in the paper. I-optimal design is also obtained in the paper and it shows that equal-weight design has high I-efficiency.

     

/

返回文章
返回