no. 3-4
Statistics
Asymptotic properties of a dimension-robust quadratic dependence measure
[Propriétés asymptotiques d'une mesure de dépendance quadratique robuste aux grandes dimensions]

Présenté par : Paul Deheuvels

Achard, Sophie1
1 Brain Mapping Unit, University of Cambridge, Downing Site, Cambridge CB2 3EB, UK
Comptes Rendus. Mathématique, Tome 346 (2008) no. 3-4, pp. 213-216.

La mesure de dépendance quadratique peut-être reliée aux mesures utilisées dans les tests d'indépendance, mais étant de plus dérivable, on peut l'utiliser dans les méthodes d'analyse en composantes indépendantes. Un noyau ajustable permet d'accélérer la convergence de l'estimateur sans pour autant affecter son biais.

The quadratic dependence measure is related to measures used in independence tests, but is derivable, thus suitable for independent component analysis. An adjustable kernel allows to accelerate the convergence of the estimator without affecting the bias.

Reçu le :
Accepté le :
Publié le :
DOI : 10.1016/j.crma.2007.10.043
Achard, Sophie 1

1 Brain Mapping Unit, University of Cambridge, Downing Site, Cambridge CB2 3EB, UK 
@article{CRMATH_2008__346_3-4_213_0,
     author = {Achard, Sophie},
     title = {Asymptotic properties of a dimension-robust quadratic dependence measure},
     journal = {Comptes Rendus. Math\'ematique},
     pages = {213--216},
     publisher = {Elsevier},
     volume = {346},
     number = {3-4},
     year = {2008},
     doi = {10.1016/j.crma.2007.10.043},
     language = {en},
     url = {http://www.numdam.org/articles/10.1016/j.crma.2007.10.043/}
}
TY  - JOUR
AU  - Achard, Sophie
TI  - Asymptotic properties of a dimension-robust quadratic dependence measure
JO  - Comptes Rendus. Mathématique
PY  - 2008
SP  - 213
EP  - 216
VL  - 346
IS  - 3-4
PB  - Elsevier
UR  - http://www.numdam.org/articles/10.1016/j.crma.2007.10.043/
DO  - 10.1016/j.crma.2007.10.043
LA  - en
ID  - CRMATH_2008__346_3-4_213_0
ER  - 
%0 Journal Article
%A Achard, Sophie
%T Asymptotic properties of a dimension-robust quadratic dependence measure
%J Comptes Rendus. Mathématique
%D 2008
%P 213-216
%V 346
%N 3-4
%I Elsevier
%U http://www.numdam.org/articles/10.1016/j.crma.2007.10.043/
%R 10.1016/j.crma.2007.10.043
%G en
%F CRMATH_2008__346_3-4_213_0
Achard, Sophie. Asymptotic properties of a dimension-robust quadratic dependence measure. Comptes Rendus. Mathématique, Tome 346 (2008) no. 3-4, pp. 213-216. doi : 10.1016/j.crma.2007.10.043. http://www.numdam.org/articles/10.1016/j.crma.2007.10.043/

[1] S. Achard, D.T. Pham, C. Jutten, Quadratic dependence measure for non linear blind sources separation, in: Proc. Int. Workshop ICA2003, April 2003, pp. 263–268

[2] Blum, J.R.; Kiefer, J.; Rosenblatt, M. Distribution free tests of independence based on the sample distribution functions, Ann. Math. Stat., Volume 32 (1961), pp. 485-498

[3] Chen, A.; Bickel, P.J. Consistent independent component analysis and prewhitening, IEEE Trans. Signal Processing, Volume 53 (2005) no. 10, pp. 3625-3632

[4] Csörgő, S. Limit behaviour of the empirical characteristic function, Ann. Probab., Volume 9 (1981) no. 1, pp. 130-144

[5] Eriksson, J.; Koivunen, V. Characteristic function based independent component analysis, Signal Processing, Volume 83 (2003), pp. 2195-2208

[6] Feuerverger, A. A consistent test for bivariate dependence, Int. Statist. Rev., Volume 61 (1993) no. 3, pp. 419-433

[7] Hoeffding, W. A class of statistics with asymptotically normal distribution, Ann. Math. Stat., Volume 19 (1948), pp. 293-325

[8] Hoeffding, W. A non-parametric test of independence, Ann. Math. Stat., Volume 19 (1948), pp. 546-557

[9] A. Kankainen, Consistent testing of total independence based on empirical characteristic functions, PhD thesis, University of Jyväskylä, 1995

[10] Rosenblatt, M. A quadratic measure of deviation of two-dimensional density estimates and a test of independence, Ann. Statist., Volume 3 (1975) no. 1, pp. 1-14

Cité par Sources :