Advertisement

Methodology and Computing in Applied Probability

, Volume 18, Issue 3, pp 911–920 | Cite as

Expansions for Log Densities of Multivariate Estimates

  • Christopher S. Withers
  • Saralees NadarajahEmail author
Article
  • 70 Downloads

Abstract

Withers and Nadarajah (Stat Pap 51:247–257; 2010) gave simple Edgeworth-type expansions for log densities of univariate estimates whose cumulants satisfy standard expansions. Here, we extend the Edgeworth-type expansions for multivariate estimates with coefficients expressed in terms of Bell polynomials. Their advantage over the usual Edgeworth expansion for the density is that the kth term is a polynomial of degree only k + 2, not 3k. Their advantage over those in Takemura and Takeuchi [Sankhyā, A, 50, 1998, 111-136] is computational efficiency

Keywords

Bell polynomials Multivariate normal distribution Normal distribution 

Mathematics Subject Classification (2010)

62E20 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. Azzalini A, Dalla Valle A (1996) The multivariate skew-normal distribution. Biometrika 83:715–726MathSciNetCrossRefzbMATHGoogle Scholar
  2. Bhattacharya RN, Ghosh JK (1978) On the validity of the formal Edgeworth expansion. Ann Stat 6:434–451MathSciNetCrossRefzbMATHGoogle Scholar
  3. Comtet L (1974) Advanced combinatorics Reidel, DordrechtGoogle Scholar
  4. Eggers HC, Lipa P (2007) Quantitative characterisation of correlation function shapes with a multivariate Edgeworth expansion. International Journal of Modern Physics E: Nuclear Physics 16:3205–3223CrossRefGoogle Scholar
  5. Eltoft T, Kim T, Lee TW (2006) On the multivariate Laplace distribution. IEEE Signal Process Lett. 13:300–303CrossRefGoogle Scholar
  6. Kim HM, Genton MG (2011) Characteristic functions of scale mixtures of multivariate skew-normal distributions. J Multivar Anal 102:1105–1117MathSciNetCrossRefzbMATHGoogle Scholar
  7. Kotz S, Nadarajah S (2004) Multivariate t distributions and their applications. Cambridge University Press, New YorkCrossRefzbMATHGoogle Scholar
  8. Perninge M (2014) Stochastic optimal power flow by multi-variate Edgeworth expansions. Electr Power Syst Res 109:90–100CrossRefGoogle Scholar
  9. Takemura A, Takeuchi K (1988) Some results on univariate and multivariate Cornish-Fisher expansions: Algebraic properties and validity. Sankhyā A 50:111–136MathSciNetzbMATHGoogle Scholar
  10. van Hulle MM (2005) Edgeworth approximation of multivariate differential entropy. Neural Comput 17:1903–1910CrossRefzbMATHGoogle Scholar
  11. Withers CS (1982) The distribution and quantiles of a function of parameter estimates. Annals of the Institute of Statistical Mathematics A 34:55–68MathSciNetCrossRefzbMATHGoogle Scholar
  12. Withers CS (1983) Expansions for the distribution and quantiles of a regular functional of the empirical distribution with applications to nonparametric confidence intervals. Ann Stat 11:577–587MathSciNetCrossRefzbMATHGoogle Scholar
  13. Withers CS (1984) Asymptotic expansions for distributions and quantiles with power series cumulants. Journal of the Royal Statistical Society B 46:389–396MathSciNetzbMATHGoogle Scholar
  14. Withers CS (2000) A simple expression for the multivariate Hermite polynomials. Statistics and Probability Letters 47:165–169MathSciNetCrossRefzbMATHGoogle Scholar
  15. Withers CS, Nadarajah S (2010) Expansions for log densities of asymptotically normal estimates. Stat Pap 51:247–257MathSciNetCrossRefzbMATHGoogle Scholar
  16. Withers CS, Nadarajah S (2014) The dual multivariate Charlier and Edgeworth expansions. Statistics and Probability Letters 87:76–85MathSciNetCrossRefzbMATHGoogle Scholar
  17. Xu H, Chen CH, Shi X (2007) Using the penalized mutual information criterion in the multivariate Edgeworth-expanded Gaussian mixture density for blind separation of convolutive post-nonlinear mixtures. Circuits Systems and Signal Processing 26:651–670MathSciNetCrossRefzbMATHGoogle Scholar

Copyright information

© Springer Science+Business Media New York 2016

Authors and Affiliations

  1. 1.Industrial Research LimitedLower HuttNew Zealand
  2. 2.University of ManchesterManchesterUK

Personalised recommendations