[Sur le volume de l'intersection d'une boule avec des demi espaces aléatoires]
Nous trouvons une expression asymptotique du volume de l'intersection d'une boule à N dimensions avec p=αN demi espaces aléatoires quand α ne depasse pas la valeur critique αc. Cette expression est la même que celle trouvée par Gardner [3] en utilisant un calcul de repliques. Nous trouvons aussi la mème valeur de αc. Notre démonstration est rigoureuse et basée sur la methode de la cavité. La nécessaire décroissance des corrélations est obtenue en utilisant un argument géométrique qui est vrai pour des hamiltoniens généraux.
We find an asymptotic expression of the volume of the intersection of the N dimensional sphere with p=αN random half spaces when α is less than a critical value. This expression coincides with the one found by Gardner [3] using replica calculations. We get also the same value for αc. Our proof is rigorous and based on the cavity method. The required decay of correlations is obtained by means of a geometrical argument which holds for general Hamiltonians.
Accepté le :
Publié le :
@article{CRMATH_2002__334_9_803_0,
author = {Maria Shcherbina and Brunello Tirozzi},
title = {On the volume of the intersection of a sphere with random half spaces},
journal = {Comptes Rendus. Math\'ematique},
pages = {803--806},
publisher = {Elsevier},
volume = {334},
number = {9},
year = {2002},
doi = {10.1016/S1631-073X(02)02345-2},
language = {en},
}
Maria Shcherbina; Brunello Tirozzi. On the volume of the intersection of a sphere with random half spaces. Comptes Rendus. Mathématique, Volume 334 (2002) no. 9, pp. 803-806. doi : 10.1016/S1631-073X(02)02345-2. https://comptes-rendus.academie-sciences.fr/mathematique/articles/10.1016/S1631-073X(02)02345-2/
[1] Hopfield models as a generalized random mean field models (A. Bovier; P. Picco, eds.), Mathematical Aspects of Spin Glasses and Neuronal Networks, Progr. Probab., 41, Birkhäuser, 1998, pp. 3-89
[2] On the extension of the Brunn–Minkowsky and Pekoda–Leindler theorems, includings inequalities for log concave functions, and with an application to the diffusion equation, J. Func. Anall., Volume 22 (1976), pp. 366-389
[3] The space of interactions in neural network models, J. Phys. A, Volume 21 (1988), pp. 271-284
[4] Optimal stage properties of neural network models, J. Phys. A, Volume 21 (1988), pp. 257-270
[5] General properties of overlap probability distributions in disordered spin system, J. Phys. A, Volume 31 (1988), pp. 9149-9155
[6] Vorlesungen über Inhalt, Oberlache und Isoperimetrie, Springer-Verlag, 1957
[7] Spin Glass Theory and Beyond, World Scientific, Singapore, 1987
[8] Absence of self-averaging of the order parameter in the Sherrington–Kirkpatrick model, J. Statist. Phys., Volume 62 (1991), pp. 1-26
[9] The replica-symmetric solution without replica trick for the hopfield model, J. Statist. Phys., Volume 74 (1994) no. 5/6, pp. 1161-1183
[10] On the replica symmetric equations for the hopfield model, J. Math. Phys., Volume 40 (1999) no. 8, pp. 3930-3947
[11] On the replica symmetric solution for the Sherrington–Kirkpatrick model, Helv. Phys. Acta, Volume 70 (1997), pp. 772-797
[12] Some estimates for the critical temperature of the Sherrington–Kirkpatrick model with magnetic field, Mathematical Results in Statistical Mechanics, World Scientific, Singapore, 1999, pp. 455-474
[13] Rigorous results for the Hopfield model with many patterns, Probab. Theory Related Fields, Volume 110 (1998), pp. 176-277
[14] Exponential inequalities and replica symmetry breaking for the Sherrington–Kirkpatrick model, Ann. Probab., Volume 28 (2000), pp. 1018-1068
[15] Intersecting random half-spaces: Toward the Gardner–Derrida problem, Ann. Probab., Volume 28 (2000), pp. 725-758
[16] Self averaging and the space of interactions in neural networks, Random Structures and Algorithms, Volume 14 (1988), pp. 199-213
Cité par Sources :
Commentaires - Politique
C. R. Math (2002)