Title :
Diffusion approximation of frequency sensitive competitive learning
Author :
Galanopoulos, Aristides S. ; Moses, Randolph L. ; Ahalt, Stanley C.
Author_Institution :
Dept. of Electr. Eng., Ohio State Univ., Columbus, OH, USA
fDate :
9/1/1997 12:00:00 AM
Abstract :
The focus of this paper is a convergence study of the frequency sensitive competitive learning (FSCL) algorithm. We approximate the final phase of FSCL learning by a diffusion process described by the Fokker-Plank equation. Sufficient and necessary conditions are presented for the convergence of the diffusion process to a local equilibrium. The analysis parallels that by Ritter-Schulten (1988) for Kohonen´s self-organizing map. We show that the convergence conditions involve only the learning rate and that they are the same as the conditions for weak convergence described previously. Our analysis thus broadens the class of algorithms that have been shown to have these types of convergence characteristics
Keywords :
approximation theory; convergence of numerical methods; diffusion; learning systems; probability; self-organising feature maps; unsupervised learning; vector quantisation; Fokker-Plank equation; convergence; diffusion approximation; frequency sensitive competitive learning; learning systems; necessary condition; probability; self-organizing map; sufficient condition; vector quantisation; Algorithm design and analysis; Convergence; Diffusion processes; Equations; Frequency; Learning systems; Neural networks; Power capacitors; Stochastic processes; Vector quantization;
Journal_Title :
Neural Networks, IEEE Transactions on