×

Multivariate generalized Gram-Charlier series in vector notations. (English) Zbl 1394.62014

Summary: This article derives the generalized Gram-Charlier (GGC) series in multivariate that expands an unknown joint probability density function (pdf) of a random vector in terms of the differentiations of joint pdf of a known reference random vector. Conventionally, the higher order differentiations of a multivariate pdf and corresponding to it the multivariate GGC series use multi-element array or tensor representations. Instead, the current article derives them in vector notations. The required higher order differentiations of a multivariate pdf are achieved in vector notations through application of a specific Kronecker product based differentiation operator. The resultant multivariate GGC series expression is more compact and more elementary compare to the coordinatewise tensor notations as using vector notations. It is also more comprehensive as apparently more nearer to its counterpart for univariate. Same notations and advantages are shared by other expressions obtained in the article, such as the mutual relations between cumulants and moments of a random vector, integral form of a multivariate pdf, integral form of the multivariate Hermite polynomials, the multivariate Gram-Charlier A series and others. Overall, the article uses only elementary calculus of several variables instead of tensor calculus to achieve the extension of a specific derivation for the GGC series in univariate [M. N. Berberan-Santos, ibid. 42, No. 3, 585–594 (2007; Zbl 1129.62007)] to multivariate.

MSC:

62E17 Approximations to statistical distributions (nonasymptotic)
62H10 Multivariate distribution of statistics
60E10 Characteristic functions; other transforms
62E20 Asymptotic distribution theory in statistics

Citations:

Zbl 1129.62007

Software:

FastICA
PDFBibTeX XMLCite
Full Text: DOI arXiv

References:

[1] S. Amari, A. Cichocki, H.H. Yang et al., in Advances in Neural Information Processing Systems. A new learning algorithm for blind signal separation (1996), pp. 757-763 · Zbl 1331.62084
[2] Amari, S; Kumon, M, Differential geometry of Edgeworth expansions in curved exponential family, Ann. Inst. Stat. Math., 35, 1-24, (1983) · Zbl 0527.62034 · doi:10.1007/BF02480959
[3] Aroian, LA, The type b Gram-Charlier series, Ann. Math. Stat., 8, 183-192, (1937) · JFM 63.1080.03 · doi:10.1214/aoms/1177732388
[4] Aroian, LA, The probability function of the product of two normally distributed variables, Ann. Math. Stat., 18, 265-271, (1947) · Zbl 0041.45004 · doi:10.1214/aoms/1177730442
[5] Berberan-Santos, M, Expressing a probability density function in terms of another pdf: a generalized Gram-Charlier expansion, J. Math. Chem., 42, 585-594, (2007) · Zbl 1129.62007 · doi:10.1007/s10910-006-9134-5
[6] Berkowitz, S; Garner, F, The calculation of multidimensional Hermite polynomials and Gram-Charlier coefficients, Math. Comput., 24, 537-545, (1970) · Zbl 0216.49301 · doi:10.1090/S0025-5718-1970-0273784-2
[7] R. Boscolo, V.P. Roychowdhury, in ica03. On the uniqueness of the minimum of the information-theoretic cost function for the separation of mixtures of nearly Gaussian signals (Nara, 2003), pp. 137-141
[8] Bowers, NL, Expansion of probability density functions as a sum of gamma densities with applications in risk theory, Trans. Soc. Actuar., 18, 125, (1966)
[9] Cohen, L, Generalization of the Gram-Charlier/Edgeworth series and application to time-frequency analysis, Multidimens. Syst. Signal Process., 9, 363-372, (1998) · Zbl 0921.60013 · doi:10.1023/A:1008454223082
[10] Cohen, L, On the generalization of the Edgeworth/Gram-Charlier series, J. Math. Chem., 49, 625-628, (2011) · Zbl 1217.62013 · doi:10.1007/s10910-010-9787-y
[11] Davis, AW, Statistical distributions in univariate and multivariate Edgeworth populations, Biometrika, 63, 661-670, (1976) · Zbl 0348.62008 · doi:10.1093/biomet/63.3.661
[12] E.B. Del Brio, T.M. Niguez, J. Perote, Gram-Charlier densities: a multivariate approach. Quant. Finance 9(7), 855-868 (2009) · Zbl 1180.91319
[13] M. Girolami, C. Fyfe, in NIPS-96 Blind Signal Separation Workshop, vol. 8, ed. by A. C. Back A. Negentropy and kurtosis as projection pursuit indices provide generalised ICA algorithms (1996) · Zbl 0843.62061
[14] Hald, A, The early history of the cumulants and the Gram-Charlier series, Int. Stat. Rev., 68, 137-153, (2000) · Zbl 1107.01304 · doi:10.1111/j.1751-5823.2000.tb00318.x
[15] A. Hald, J. Steffensen, On the History of Series Expansions of Frequency Functions and Sampling Distributions, 1873-1944 (Det Kongelige Danske Videnskabernes Selskab, Copenhagen, 2002)
[16] Holmquist, B, The d-variate vector Hermite polynomial of order k, Linear Algebra Appl., 237-238, 155-190, (1996) · Zbl 0848.62027 · doi:10.1016/0024-3795(95)00595-1
[17] A. Hyvärinen, J. Karhunen, E. Oja, Independent Component Analysis (Wiley, New York, 2001), p. 481+xxii · doi:10.1002/0471221317
[18] Jammalamadaka, SR; Rao, TS; Terdik, G, Higher order cumulants of random vectors and applications to statistical inference and time series, Sankhya Indian J. Stat. (2003-2007), 68, 326-356, (2006) · Zbl 1193.62097
[19] Jondeau, E; Rockinger, M, Gram-Charlier densities, J. Econ. Dyn. Control, 25, 1457-1483, (2001) · Zbl 1056.91512 · doi:10.1016/S0165-1889(99)00082-2
[20] Kollo, T; Rosen, D, Approximating by the Wishart distribution, Ann. Inst. Stat. Math., 47, 767-783, (1995) · Zbl 0843.62061 · doi:10.1007/BF01856546
[21] Kollo, T; Rosen, D, A unified approach to the approximation of multivariate densities, Scand. J. Stat., 25, 93-109, (1998) · Zbl 0903.62051 · doi:10.1111/1467-9469.t01-1-00091
[22] Magnus, JR, On the concept of matrix derivative, J. Multivar. Anal., 101, 2200-2206, (2010) · Zbl 1196.15025 · doi:10.1016/j.jmva.2010.05.005
[23] J.R. Magnus, H. Neudecker, Matrix Differential Calculus with Applications in Statistics and Econometrics (Wiley, Chichester, 1999) · Zbl 0912.15003
[24] P. McCullagh, Tensor Methods in Statistics (Chapman & Hall, London, 1987) · Zbl 0732.62003
[25] M.C.M. O’Brien, Using the Gram-Charlier expansion to produce vibronic band shapes in strong coupling. J. Phys. Condens. Matter 4(9), 2347-2359 (1992). http://stacks.iop.org/0953-8984/4/i=9/a=027
[26] H.L. Rietz, in Carus Mathematical Monographs, vol. 3, 1st edn. Mathematical Statistics (Mathematical Association of America, Washington, 1927)
[27] Sauer, P; Heydt, G, A convenient multivariate Gram-Charlier type a series, IEEE Trans. Commun., 27, 247-248, (1979) · Zbl 0399.62015 · doi:10.1109/TCOM.1979.1094247
[28] Schleher, DC, Generalized Gram-Charlier series with application to the sum of log-normal variates (corresp.), IEEE Trans. Inf. Theory, 23, 275-280, (1977) · Zbl 0358.60023 · doi:10.1109/TIT.1977.1055686
[29] Skovgaard, IM, On multivariate Edgeworth expansions, Int. Stat. Rev., 54, 169-186, (1986) · Zbl 0615.62018 · doi:10.2307/1403142
[30] Terdik, G, Higher order statistics and multivariate vector Hermite polynomials, Teor. Imovir. Mat. Stat., 66, 147-168, (2002) · Zbl 1026.62090
[31] Viehland, LA, Velocity distribution functions and transport coefficients of atomic ions in atomic gases by a Gram-Charlier approach, Chem. Phys., 179, 71-92, (1994) · doi:10.1016/0301-0104(93)E0337-U
[32] Withers, CS; Nadarajah, S, The dual multivariate Charlier and Edgeworth expansions, Stat. Probab. Lett., 87, 76-85, (2014) · Zbl 1331.62084 · doi:10.1016/j.spl.2014.01.003
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.