Matrix variate generalized asymmetric Laplace distributions

Kozubowski, Tomasz; Mazur, Stepan; Podgórski, Krzysztof

doi:10.1090/tpms/1197

Authors: Tomasz J. Kozubowski, Stepan Mazur and Krzysztof Podgórski
Journal: Theor. Probability and Math. Statist. 109 (2023), 55-80
MSC (2020): Primary 62H10, 60E05; Secondary 60E10
DOI: https://doi.org/10.1090/tpms/1197
Published electronically: October 3, 2023
Full-text PDF

Abstract | References | Similar Articles | Additional Information

Abstract: The generalized asymmetric Laplace (GAL) distributions, also known as the variance/mean-gamma models, constitute a popular flexible class of distributions that can account for peakedness, skewness, and heavier-than-normal tails, often observed in financial or other empirical data. We consider extensions of the GAL distribution to the matrix variate case, which arise as covariance mixtures of matrix variate normal distributions. Two different mixing mechanisms connected with the nature of the random scaling matrix are considered, leading to what we term matrix variate GAL distributions of Type I and II. While Type I matrix variate GAL distribution has been studied before, there is no comprehensive account of Type II in the literature, except for their rather brief treatment as a special case of matrix variate generalized hyperbolic distributions. With this work we fill this gap, and present an account for basic distributional properties of Type II matrix variate GAL distributions. In particular, we derive their probability density function and the characteristic function, as well as provide stochastic representations related to matrix variate gamma distribution. We also show that this distribution is closed under linear transformations, and study the relevant marginal distributions. In addition, we also briefly account for Type I and discuss the intriguing connections with Type II. We hope that this work will be useful in the areas where matrix variate distributions provide an appropriate probabilistic tool for three-way or, more generally, panel data sets, which can arise across different applications.

References

Laura Anderlucci and Cinzia Viroli, Covariance pattern mixture models for the analysis of multivariate heterogeneous longitudinal data, Ann. Appl. Stat. 9 (2015), no. 2, 777–800. MR 3371335, DOI 10.1214/15-AOAS816
Anis Iranmanesh, M. Arashi, and S. M. M. Tabatabaey, On conditional applications of matrix variate normal distribution, Iran. J. Math. Sci. Inform. 5 (2010), no. 2, 33–43 (English, with English and Persian summaries). MR 2866385
H. Asgharian, K. Podgórski, F. N. Shariati, and L. Liu, Structural Multivariate Spatial Econometrics: Application to Cross-Country Interdependence of Stock and Bond markets, Available at SSRN: https://ssrn.com/abstract=3527346, 2018.
O. Barndorff-Nielsen, J. Kent, and M. Sørensen, Normal variance-mean mixtures and $z$ distributions, Internat. Statist. Rev. 50 (1982), no. 2, 145–159 (English, with French summary). MR 678296, DOI 10.2307/1402598
Ole E. Barndorff-Nielsen, Víctor Pérez-Abreu, and Alfonso Rocha-Arteaga, $\textrm {mat}\,G$ random matrices, Stoch. Models 22 (2006), no. 4, 723–734. MR 2263863, DOI 10.1080/15326340600878560
A. Bekker and J. J. J. Roux, Bayesian multivariate normal analysis with a Wishart prior, Comm. Statist. Theory Methods 24 (1995), no. 10, 2485–2497. MR 1354468, DOI 10.1080/03610929508831629
T. Bodnar, S. Mazur, S. Muhinyuza, and N. Parolya, On the product of a singular Wishart matrix and a singular Gaussian vector in high dimension, Teor. Ĭmovīr. Mat. Stat. 99 (2018), 37–50 (English, with English, Russian and Ukrainian summaries); English transl., Theory Probab. Math. Statist. 99 (2019), 39–52. MR 3908654, DOI 10.1090/tpms/1078
T. Bodnar, S. Mazur, and Y. Okhrin, Distribution of the product of singular Wishart matrix and normal vector, Teor. Ĭmovīr. Mat. Stat. 91 (2014), 1–14; English transl., Theory Probab. Math. Statist. 91 (2015), 1–15. MR 3364119, DOI 10.1090/tpms/962
Taras Bodnar, Stepan Mazur, and Krzysztof Podgórski, Singular inverse Wishart distribution and its application to portfolio theory, J. Multivariate Anal. 143 (2016), 314–326. MR 3431434, DOI 10.1016/j.jmva.2015.09.021
Y. Murat Bulut and Olcay Arslan, Matrix variate slash distribution, J. Multivariate Anal. 137 (2015), 173–178. MR 3332805, DOI 10.1016/j.jmva.2015.02.008
A. P. Dawid, Some matrix-variate distribution theory: notational considerations and a Bayesian application, Biometrika 68 (1981), no. 1, 265–274. MR 614963, DOI 10.1093/biomet/68.1.265
James M. Dickey, Matricvariate generalizations of the multivariate $t$ distribution and the inverted multivariate $t$ distribution, Ann. Math. Statist. 38 (1967), 511–518. MR 208752, DOI 10.1214/aoms/1177698967
Jiu Ding and Aihui Zhou, Eigenvalues of rank-one updated matrices with some applications, Appl. Math. Lett. 20 (2007), no. 12, 1223–1226. MR 2384251, DOI 10.1016/j.aml.2006.11.016
Peng Ding and Joseph K. Blitzstein, On the Gaussian mixture representation of the Laplace distribution, Amer. Statist. 72 (2018), no. 2, 172–174. MR 3810626, DOI 10.1080/00031305.2017.1291448
F. Z. Doğru, Y. M. Bulut, and O. Arslan, Finite mixtures of matrix variate t distributions, Gazi University Journal of Science 29 (2016), no. 2, 335–341.
Michael P. B. Gallaugher and Paul D. McNicholas, A matrix variate skew-$t$ distribution, Stat 6 (2017), 160–170. MR 3653050, DOI 10.1002/sta4.143
—, Finite mixtures of skewed matrix variate distributions, Pattern Recognition 80 (2018), 83–93.
Michael P. B. Gallaugher and Paul D. McNicholas, Three skewed matrix variate distributions, Statist. Probab. Lett. 145 (2019), 103–109. MR 3873895, DOI 10.1016/j.spl.2018.08.012
Michael P. B. Gallaugher and Paul D. McNicholas, Mixtures of skewed matrix variate bilinear factor analyzers, Adv. Data Anal. Classif. 14 (2020), no. 2, 415–434. MR 4118954, DOI 10.1007/s11634-019-00377-4
Seymour Geisser, Bayesian estimation in multivariate analysis, Ann. Math. Statist. 36 (1965), 150–159. MR 174134, DOI 10.1214/aoms/1177700279
A. K. Gupta and D. K. Nagar, Matrix variate distributions, Chapman & Hall/CRC Monographs and Surveys in Pure and Applied Mathematics, vol. 104, Chapman & Hall/CRC, Boca Raton, FL, 2000. MR 1738933
A. K. Gupta and T. Varga, Normal mixture representations of matrix variate elliptically contoured distributions, Sankhyā Ser. A 57 (1995), no. 1, 68–78. MR 1392632
Solomon W. Harrar, Eugene Seneta, and Arjun K. Gupta, Duality between matrix variate $t$ and matrix variate V.G. distributions, J. Multivariate Anal. 97 (2006), no. 6, 1467–1475. MR 2256162, DOI 10.1016/j.jmva.2005.09.002
David A. Harville, Matrix algebra from a statistician’s perspective, Springer-Verlag, New York, 1997. MR 1467237, DOI 10.1007/b98818
Carl S. Herz, Bessel functions of matrix argument, Ann. of Math. (2) 61 (1955), 474–523. MR 69960, DOI 10.2307/1969810
Md. Mobarak Hossain, Tomasz J. Kozubowski, and Krzysztof Podgórski, A novel weighted likelihood estimation with empirical Bayes flavor, Comm. Statist. Simulation Comput. 47 (2018), no. 2, 392–412. MR 3757694, DOI 10.1080/03610918.2016.1197246
B. M. Golam Kibria, The matrix-$t$ distribution and its applications in predictive inference, J. Multivariate Anal. 97 (2006), no. 3, 785–795. MR 2236502, DOI 10.1016/j.jmva.2005.08.001
Samuel Kotz, Tomasz J. Kozubowski, and Krzysztof Podgórski, The Laplace distribution and generalizations, Birkhäuser Boston, Inc., Boston, MA, 2001. A revisit with applications to communications, economics, engineering, and finance. MR 1935481, DOI 10.1007/978-1-4612-0173-1
T. J. Kozubowski, S. Mazur, and K. Podgórski, Matrix gamma distributions and related stochastic processes, Working Paper 12, Örebro University, 2022.
Tomasz J. Kozubowski and Krzysztof Podgórski, Gaussian mixture representation of the Laplace distribution revisited: bibliographical connections and extensions, Amer. Statist. 74 (2020), no. 4, 407–412. MR 4168260, DOI 10.1080/00031305.2019.1630000
Tomasz J. Kozubowski, Krzysztof Podgórski, and Igor Rychlik, Multivariate generalized Laplace distribution and related random fields, J. Multivariate Anal. 113 (2013), 59–72. MR 2984356, DOI 10.1016/j.jmva.2012.02.010
D. B. Madan and E. Seneta, The variance gamma (V.G.) model for share market returns, The Journal of Business 63 (1990), no. 4, 511–524.
Pierre-Alexandre Mattei, Multiplying a Gaussian matrix by a Gaussian vector, Statist. Probab. Lett. 128 (2017), 67–70. MR 3656377, DOI 10.1016/j.spl.2017.04.004
Volodymyr Melnykov and Xuwen Zhu, On model-based clustering of skewed matrix data, J. Multivariate Anal. 167 (2018), 181–194. MR 3830641, DOI 10.1016/j.jmva.2018.04.007
Anne Opschoor, Pawel Janus, André Lucas, and Dick Van Dijk, New HEAVY models for fat-tailed realized covariances and returns, J. Bus. Econom. Statist. 36 (2018), no. 4, 643–657. MR 3871707, DOI 10.1080/07350015.2016.1245622
Victor Pérez-Abreu and Robert Stelzer, Infinitely divisible multivariate and matrix gamma distributions, J. Multivariate Anal. 130 (2014), 155–175. MR 3229530, DOI 10.1016/j.jmva.2014.04.017
L. Thabane and M. Safiul Haq, On the matrix-variate generalized hyperbolic distribution and its Bayesian applications, Statistics 38 (2004), no. 6, 511–526. MR 2109632, DOI 10.1080/02331880412331319279
Geoffrey Z. Thompson, Ranjan Maitra, William Q. Meeker, and Ashraf F. Bastawros, Classification with the matrix-variate-$t$ distribution, J. Comput. Graph. Statist. 29 (2020), no. 3, 668–674. MR 4153190, DOI 10.1080/10618600.2019.1696208
Cinzia Viroli, Finite mixtures of matrix normal distributions for classifying three-way data, Stat. Comput. 21 (2011), no. 4, 511–522. MR 2826689, DOI 10.1007/s11222-010-9188-x
Cinzia Viroli, Model based clustering for three-way data structures, Bayesian Anal. 6 (2011), no. 4, 573–602. MR 2869958, DOI 10.1214/11-BA622
Y. Yurchenko, Matrix variate and tensor variate Laplace distributions, Available at arXiv: https://arxiv.org/abs/2104.05669, 2021.