Asymptotic behavior of self-organizing maps with nonuniform stimuli distribution
Sadeghi, Ali A.
Ann. Appl. Probab., Tome 8 (1998) no. 1, p. 281-299 / Harvested from Project Euclid
Here the almost sure convergence of one-dimensional Kohonen's algorithm in its general form, namely, the 2k-neighbor setting with a nonuniform stimuli distribution, is proved. We show that the asymptotic behavior of the algorithm is governed by a cooperative system of differential equations which is irreducible. The system of differential equations possesses an asymptotically stable equilibrium, a compact subset of whose domain of attraction will be visited by the state variable $X^n$ infinitely often. The assumptions on the stimuli distribution and the neighborhood functions are weakened, too.
Publié le : 1998-02-14
Classification:  Neural networks,  stochastic approximation,  theory of differential equations,  60J05,  93D20,  92B20
@article{1027961044,
     author = {Sadeghi, Ali A.},
     title = {Asymptotic behavior of self-organizing maps with nonuniform
		 stimuli distribution},
     journal = {Ann. Appl. Probab.},
     volume = {8},
     number = {1},
     year = {1998},
     pages = { 281-299},
     language = {en},
     url = {http://dml.mathdoc.fr/item/1027961044}
}
Sadeghi, Ali A. Asymptotic behavior of self-organizing maps with nonuniform
		 stimuli distribution. Ann. Appl. Probab., Tome 8 (1998) no. 1, pp.  281-299. http://gdmltest.u-ga.fr/item/1027961044/