Self similar compound symmetry covariance structure. (English) Zbl 1477.62129

Leiva, Ricardo; Roy, Anuradha

J. Stat. Theory Pract. 15, No. 3, Paper No. 70, 31 p. (2021).

Summary: Self similar compound symmetry (SSCS) covariance structure is introduced and studied. The \(k\)-SSCS covariance structure (defined in Sect. 3) for array-variate \(k\)th order data incorporates the exchangeable feature of \(k\)-dimensional arrays into the model. 3-SSCS covariance structure or double block compound symmetry covariance structure for array-variate 3rd order data is a generalization of 2-SSCS covariance structure or block compound symmetry covariance structure for the matrix-variate 2nd order data, which in turn is a generalization of compound symmetry covariance structure for traditional vector-variate (multivariate) 1st order data. This article generalizes this compound symmetry covariance structure for array-variate \(k\)th order data, and we name it as “\(k\) self similar compound symmetry” (\(k\)SSCS) covariance structure. This is of critical importance to a variety of applied problems in agricultural, biomedical, medical, environmental, engineering and space missions among many other fields with \(k\)-dimensional array-variate data. The proposed method is illustrated with a medical dataset.

MSC:

62H10	Multivariate distribution of statistics
62H12	Estimation in multivariate analysis
62P10	Applications of statistics to biology and medical sciences; meta analysis

Keywords:

array-variate data; eigenblock; high dimensional data; Wishart distribution

References:

[1]	Anderson, TW, An introduction to multivariate statistical analysis (2003), Hoboken: Wiley, Hoboken · Zbl 1039.62044
[2]	Arnold, SF, Application of the theory of products of problems to certain patterned covariance matrices, Ann Stat, 1, 4, 682-699 (1973) · Zbl 0289.62038
[3]	Coelho, CA; Roy, A., Testing the hypothesis of a doubly exchangeable covariance matrix, Metrika, 83, 1, 45-68 (2020) · Zbl 1435.62198
[4]	Hao, C.; Liang, Y.; Roy, A., Equivalency between vertices and centers-coupled-with-radii principal component analyses for interval data, Stat Probabil Lett, 106, 113-120 (2015) · Zbl 1328.62374
[5]	Johnson, RA; Wichern, DW, Applied multivariate statistical analysis (2007), Englewood Cliffs: Pearson Prentice Hall, Englewood Cliffs · Zbl 1269.62044
[6]	Klein, D.; Žežula, I., Orthogonal decompositions in growth curve models, Acta et Commentationes Universitatis Tartuensis de Mathematica, 14, 35-44 (2010) · Zbl 1228.62065
[7]	Leiva, R., Linear discrimination with equicorrelated training vectors, J Multivar Anal, 98, 384-409 (2007) · Zbl 1112.62057
[8]	Leiva, R.; Roy, A., Linear discrimination for three-level multivariate data with separable additive mean vector and doubly exchangeable covariance structure, Comput Stat Data Anal, 56, 6, 1644-1661 (2012) · Zbl 1243.62094
[9]	Liang, L.; von Rosen, D.; von Rosen, T., On estimation in hierarchical models with block circular covariance structures, Ann Inst Stat Math, 67, 4, 773-791 (2015) · Zbl 1440.62273
[10]	Mardia, KV; Kent, JT; Bibby, JM, Multivariate analysis (1979), New York: Academic Press Inc., New York · Zbl 0432.62029
[11]	Manceur, AM; Dutilleul, P., Maximum likelihood estimation for the tensor normal distribution: algorithm, minimum sample size, and empirical bias and dispersion, J Comput Appl Math, 239, 37-49 (2013) · Zbl 1255.65029
[12]	Nahtman, T., Marginal permutation invariant covariance matrices with applications to linear models, Linear Algebra Appl, 417, 183-210 (2006) · Zbl 1099.62074
[13]	Ohlson, M.; Ahmad, MR; von Rosen, D., The multilinear normal distribution: introduction and some basic properties, J Multivar Anal, 113, 37-47 (2013) · Zbl 1354.60015
[14]	Olkin, I.; Press, S., Testing and estimation for a circular stationary model, Ann Math Stat, 40, 1358-1373 (1969) · Zbl 0186.51801
[15]	Opheim T, Roy A (2019) Revisiting the linear models with exchangeably distributed errors. In: Proceedings of the American statistical association, pp 2677-2686
[16]	Opheim, T.; Roy, A., Linear models for multivariate repeated measures data with block exchangeable covariance structure, Comput Stat (2021) · Zbl 1505.62306
[17]	Opheim, T.; Roy, A., Score tests for intercept and slope parameters of doubly multivariate linear models with skew-normal errors, J Stat Theory Pract, 15, 2, 30 (2021) · Zbl 1473.62242
[18]	Rao, CR, Familial correlations or the multivariate generalizations of the intraclass correlation, Curr Sci, 14, 66-67 (1945)
[19]	Rao, CR, Discriminant functions for genetic differentiation and selection, Sankhya, 12, 229-246 (1953) · Zbl 0051.11002
[20]	Roy, A.; Leiva, R., Discrimination with jointly equicorrelated multi-level multivariate data, Adv Data Anal Classif, 1, 3, 175-199 (2007) · Zbl 1182.62142
[21]	Roy, A.; Fonseca, M., Linear models with doubly exchangeable distributed errors, Commun Stat Theory Methods, 41, 13, 2545-2569 (2012) · Zbl 1270.62102
[22]	Roy, A.; Leiva, R.; Žežula, I.; Klein, D., Testing of equality of mean vectors for paired doubly multivariate observations in blocked compound symmetric covariance matrix setup, J Multivar Anal, 137, 50-60 (2015) · Zbl 1329.62251
[23]	Szatrowski TH (1976) Estimation and testing for block compound symmetry and other patterned covariance matrices with linear and nonlinear structure, Technical Report, Number OLK NSF 107. Standford University, Department of Statistics
[24]	Žežula, I.; Klein, D.; Roy, A., Testing of multivariate repeated measures data with block exchangeable covariance structure, TEST, 27, 2, 360-378 (2018) · Zbl 1404.62062
[25]	Žežula, I.; Klein, D.; Roy, A.; Holgersson, T.; Singull, M., Mean value test for three-level multivariate observations with doubly exchangeable covariance structure, Recent developments in multivariate and random matrix analysis, 335-349 (2020), Cham: Springer, Cham