A Note on the Asymptotic Distributions of Dimensionality Estimators in Discriminant Analysis

  • Changha Hwang (Department of Statistics, CATHOLIC UNIVERSITY OF TAEGU-HYOSUNG, Kyungbuk, 713-702, KOREA.) ;
  • Daehak Kim (Department of Statistics, CATHOLIC UNIVERSITY OF TAEGU-HYOSUNG, Kyungbuk, 713-702, KOREA.)
  • Published : 1995.12.01

Abstract

The purpose of this paper is to simplify and study the asymptotic distribution of the dimensionality estimators in discriminant analysis based on Akaike's and Mallows' methods for samples from a multivariate distribution with finite fourth moments.

Keywords

References

  1. Communications in Statistics-Theory and Methods v.11 Tests of dimensionality in multivariate analysis of variance Backhouse,A.R.;McKay,R.J.
  2. Statistical Theory and Data Analysis Two methods for estimation of dimensionality in canonical correlation analysis and the multivariate linear model Fujikoshi,Y.;K.Matusita(ed.)
  3. Biomatrika v.66 Estimation of dimensionality in canonical correlation analysis Fujikoshi,Y.;Veutch,L.G.
  4. Ph.D. Thesis Estimating the dimensionality in canonical correlation analysis Gumderson,B.K.
  5. Ph.D. Thesis Model Selection Methods in Discrimination Analysis Hwang,C.
  6. Journal of Korean Statistical Society v.23 Characterization of the asymptotic distributions of certain eigenvalues setting Hwang,C.
  7. Aspects of Multivatiate Statistical Theory Muirhead,R.J.
  8. Biomatrika v.67 Asymptotic distributions in canonical correlation analysis and other multivariate procedures for nonnormal populations Muirhead,R.J.;Waternaux,C.M.
  9. Modern Multivariate Statistical Analysis: A Grafuate Course and Handbook Siotani,M.;Hayakawa,T.;Fujikoshi,Y.
  10. Journal of Multi. Anal. v.6 Asymptotic expansions of the distributions of the latent roots and the latent vectors of the Wishart and multivariate F matrices Sugiura,N.