Stability analysis of a neural field self-organizing map (Q2040386)

This paper analyses conditions that a self organising map represents an input space by using a neural field model of spatiotemporal activity in the brain and stability analysis via Lyapunov theory. These self organising maps are neural networks that map high dimensional spaces to low dimensional ones. The stability analysis provides a mechanism to identify parameter sets within which unsupervised learning is effective. This analysis is confirmed by simulations.