High-dimensional general linear hypothesis tests via non-linear spectral shrinkage

被引:6
作者
Li, Haoran [1 ]
Aue, Alexander [2 ]
Paul, Debashis [2 ]
机构
[1] Columbia Univ, Dept Stat, 1255 Amsterdam Ave, New York, NY 10027 USA
[2] Univ Calif Davis, Dept Stat, One Shields Ave, Davis, CA 95616 USA
关键词
general linear hypothesis; local alternatives; random matrix theory; ridge shrinkage; spectral shrinkage; 2-SAMPLE TEST; STATISTICS; LIMIT; EIGENVALUE; CLT;
D O I
10.3150/19-BEJ1186
中图分类号
O21 [概率论与数理统计]; C8 [统计学];
学科分类号
020208 ; 070103 ; 0714 ;
摘要
We are interested in testing general linear hypotheses in a high-dimensional multivariate linear regression model. The framework includes many well-studied problems such as two-sample tests for equality of population means, MANOVA and others as special cases. A family of rotation-invariant tests is proposed that involves a flexible spectral shrinkage scheme applied to the sample error covariance matrix. The asymptotic normality of the test statistic under the null hypothesis is derived in the setting where dimensionality is comparable to sample sizes, assuming the existence of certain moments for the observations. The asymptotic power of the proposed test is studied under various local alternatives. The power characteristics are then utilized to propose a data-driven selection of the spectral shrinkage function. As an illustration of the general theory, we construct a family of tests involving ridge-type regularization and suggest possible extensions to more complex regularizers. A simulation study is carried out to examine the numerical performance of the proposed tests.
引用
收藏
页码:2541 / 2571
页数:31
相关论文
共 36 条
[1]  
Anderson TW, 1958, Wiley Publications in Statistics
[2]  
[Anonymous], 2014, PREPRINT
[3]  
Bai ZD, 2004, ANN PROBAB, V32, P553
[4]  
Bai ZD, 1996, STAT SINICA, V6, P311
[5]   LIMITING BEHAVIOR OF EIGENVALUES IN HIGH-DIMENSIONAL MANOVA VIA RMT [J].
Bai, Zhidong ;
Choi, Kwok Pui ;
Fujikoshi, Yasunori .
ANNALS OF STATISTICS, 2018, 46 (06) :2985-3013
[6]   Testing linear hypotheses in high-dimensional regressions [J].
Bai, Zhidong ;
Jiang, Dandan ;
Yao, Jian-feng ;
Zheng, Shurong .
STATISTICS, 2013, 47 (06) :1207-1223
[7]   ON ESTIMATION OF THE POPULATION SPECTRAL DISTRIBUTION FROM A HIGH-DIMENSIONAL SAMPLE COVARIANCE MATRIX [J].
Bai, Zhidong ;
Chen, Jiaqi ;
Yao, Jianfeng .
AUSTRALIAN & NEW ZEALAND JOURNAL OF STATISTICS, 2010, 52 (04) :423-437
[8]  
Billingsley P., 2013, CONVERGE PROBAB MEAS
[9]   A nonparametric two-sample test applicable to high dimensional data [J].
Biswas, Munmun ;
Ghosh, Anil K. .
JOURNAL OF MULTIVARIATE ANALYSIS, 2014, 123 :160-171
[10]   TESTING FOR INDEPENDENCE OF LARGE DIMENSIONAL VECTORS [J].
Bodnar, Taras ;
Dette, Holger ;
Parolya, Nestor .
ANNALS OF STATISTICS, 2019, 47 (05) :2977-3008