Neural network chaos analysis

Abstract

The analysis of a chaos theory, in general, and a neural network chaos paradigm with a concrete interpretation by a simple neural network, in particular, allowed us to set up a new aspect in an arti cial neuronal approach of methodology: an analogy between natural chaos experimentally observed in neural systems of the brain and arti cial neural network chaos phenomena has been considered. The signi cance of asymmetry and nonlinearity, which were increased on introducing a restricted N-shaped synaptical relation in a two-mod dynamic model, is emphasized. There are illustrated the diferent computational examples of the neural network properties which are expressed by the equillibrium point, stable cycle or chaotic behaviour in strong nonlinear neural systems.

Keywords

• Neural network
• dynamics
• chaos theory
• chaotic computer
• equillibrium state
• symmetry