A simultaneous testing of the mean vector and the covariance matrix among two populations for high-dimensional data
A simultaneous testing of the mean vector and the covariance matrix among two populations for high-dimensional data
复制标题
DOI：
10.1007/s11749-017-0567-x
复制
发表时间：
2018-09
期刊：
影响因子：
1.3
通讯作者：
Masashi Hyodo;T. Nishiyama
中科院分区：
文献类型：
--
作者：
Masashi Hyodo;T. Nishiyama
在这篇文章中，我们提出了一个基于范数的测试，同时测试的均值向量和协方差矩阵在高维非正态总体。为了构建这个，我们推导出一个渐近分布的检验统计量的基础上的差异均值向量和协方差矩阵。我们还调查了渐近的大小和权力的建议测试使用这个结果。最后，我们研究了有限样本和维数的性能，通过Monte Carlo模拟。
In this article, we propose an-norm-based test for simultaneous testing of the mean vector and the covariance matrix under high-dimensional non-normal populations. To construct this, we derive an asymptotic distribution of a test statistic based on both differences mean vectors and covariance matrices. We also investigate the asymptotic sizes and powers of the proposed test using this result. Finally, we study the finite sample and dimension performance of this test via Monte Carlo simulations.