Informatics and Applications
2014, Volume 8, Issue 4, pp 1119
A MODIFIED GRID METHOD FOR STATISTICAL SEPARATION OF NORMAL VARIANCEMEAN MIXTURES
 V. Yu. Korolev
 A. Yu. Korchagin
Abstract
A modified twostage grid method for statistical separation of normal variancemean mixtures is described as an alternative to a pure EM (expectationmaximization) algorithm. At the first stage of this algorithm, a discrete approximation is constructed to the mixing distribution. At the second stage, the obtained discrete distribution is approximated by an absolutely continuous distribution from a predetermined family, say, by a generalized inverse Gaussian distribution. The convergence of this twostage procedure is discussed. The monotonicity of the grid procedure used at the first stage is proved. The problem of the optimal choice of the parameters of the method is discussed in detail. First of all, the problem of the optimal choice of the grid thrown on the support of the mixing distribution is considered. Statistical estimators are proposed for the quantiles of the mixing law. The efficiency of the method is illustrated by examples of its application to the estimation of the parameters of generalized hyperbolic distributions.
[+] References (12)
 Korolev, V. Yu. 2011. Veroyatnostnostatisticheskie metody dekompozitsii volatil'nosti khaoticheskikh protsessov [Probabilistic and statistical methods for the decomposition of volatility of chaotic processes]. Moscow: Moscow University Press. 510 p.
 Nazarov, A. L. 2013. Priblizhennye metody razde leniya smesey veroyatnostnykh raspredeleniy [Approximate methods forthe decomposition ofvolatility of chaotic processes]. Ph.D. Thesis. Moscow: Moscow State University.
 BarndorffNielsen, O.E. 1977. Exponentially decreasing distributions for the logarithm of particle size. Proc. Roy. Soc. Lond.A 353:401419.
 BarndorffNielsen, O.E. 1978. Hyperbolic distributions and distributions of hyperbolae. Scand. J. Statist. 5:151157.
 BarndorffNielsen, O. E., J. Kent, andM. S0rensen. 1982. Normal variancemean mixtures and zdistributions. Int. Statist. Rev. 50(2):145159.
 Korolev, V. Yu., and I. A. Sokolov. 2012. Skoshennye raspredeleniya St'yudenta, dispersionnye gammaraspre deleniya i ikh obobshcheniya kak asimptoticheskie ap proksimatsii [Skewed Student's distributions, variance
gamma distributions, and their generalizations as asymptotic approximations]. Informatika i ee Primeneniya  Inform. Appl. 6(1):210.
 Korolev, V. Yu., and L. M. Zaks. 2013. Obobshchennye dispersionnye gammaraspredeleniya kak predel'nye dlya sluchaynykh summ [Generalized variance gamma distributions as limiting for random sums]. Informatika i ee Primeneniya  Inform. Appl. 7(1):105115.
 Korolev, V. Yu. 2013. Obobshchennye giperbolicheskie raspredeleniya kak predel'nye dlya sluchaynykh summ [Generalized hyperbolic distributions as limiting for ran domsums] Theory Probab. Appl. 58(1):117132.
 Korolev, V. Yu., A. V. Chertok, A. Yu. Korchagin, and A. K. Gorshenin. 2013. Veroyatnostnostatisticheskoe modelirovanie informatsionnykh potokov v slozhnykh fi nansovykh sistemakh na osnove vysokochastotnykh dan nykh [Probability and statistical modeling ofinformation flows in complex financial systems from highfrequency data]. Informatika i ee Primeneniya  Inform. Appl. 7(1):1221.
 Protassov, R. S. 2004. EMbased maximum likelihood parameter estimation for a multivariate generalized hyperbolic distribution with fixed Ë. Statistics Computing 14:6777.
 Korolev, V. Yu., and A. L. Nazarov. 2010. Razdelenie smesey veroyatnostnykh raspredeleniy pri pomoshchi se tochnykh metodov momentov i maksimal'nogo prav dopodobiya [Separation of mixtures using grid moment based methods and maximum likelihood]. Avtomatika
i Telemekhanika [Automatics and Telemechanics] 3:98116.
 Dennis, J. E., and R. B. Schnabel. 1983. Numerical methods for unconstrained optimization and nonlinear equations. Englewood Cliffs: PrenticeHall. 378 p.
[+] About this article
Title
A MODIFIED GRID METHOD FOR STATISTICAL SEPARATION OF NORMAL VARIANCEMEAN MIXTURES
Journal
Informatics and Applications
2014, Volume 8, Issue 4, pp 1119
Cover Date
20141030
DOI
10.14357/19922264140402
Print ISSN
19922264
Publisher
Institute of Informatics Problems, Russian Academy of Sciences
Additional Links
Key words
mixture of probability distributions; normal variancemean mixture; generalized hyperbolic distribution; EMalgorithm; grid method of separation of mixtures
Authors
V. Yu. Korolev , and A. Yu. Korchagin1
Author Affiliations
Faculty of Computational Mathematics and Cybernetics, M.V. Lomonosov Moscow State University, 152 Leninskiye Gory, GSP1, Moscow 119991, Russian Federation
Federation
Institute of Informatics Problems, Russian Academy of Sciences, 442 Vavilov Str., Moscow 119333, Russian
