Skip to Main content Skip to Navigation
Journal articles

Sur la normalité asymptotique des valeurs propres en ACM sous l'hypothèse d'indépendance des variables.

Salwa Benammou 1 Gilbert Saporta 2
2 CEDRIC - MSDMA - CEDRIC. Méthodes statistiques de data-mining et apprentissage
CEDRIC - Centre d'études et de recherche en informatique et communications
Abstract : Under independence hypothesis, the eigenvalues in MCA are theoretically identical. In practice we observe that the eigenvalues are different, but close to the theoretical value. This paper deals with the distributions of these eigenvalues. We show that eigenvalues in MCA are asymptotically normal. Convergence is very slow especially for the first and the last ones. we propose an empirical procedure fot the selection of number of eigenvalues in MCA.
Document type :
Journal articles
Complete list of metadatas

Cited literature [14 references]  Display  Hide  Download

https://hal-cnam.archives-ouvertes.fr/hal-02507921
Contributor : Gilbert Saporta <>
Submitted on : Monday, March 16, 2020 - 10:49:52 AM
Last modification on : Thursday, April 2, 2020 - 4:34:20 PM
Long-term archiving on: : Wednesday, June 17, 2020 - 12:37:02 PM

File

RSA_1998__46_3_21_0.pdf
Publisher files allowed on an open archive

Identifiers

  • HAL Id : hal-02507921, version 1

Collections

Citation

Salwa Benammou, Gilbert Saporta. Sur la normalité asymptotique des valeurs propres en ACM sous l'hypothèse d'indépendance des variables.. Revue de Statistique Appliquée, Société française de statistique, 1998, 46 (3), pp.21-35. ⟨hal-02507921⟩

Share

Metrics

Record views

45

Files downloads

29