Intermittency in a Neural Network with Variable Threshold

Abstract

A model of neural network with a time-dependent threshold is presented. Changes in a control parameter of the threshold function lead the network activity to intermittent or chaotic behaviour. The intermittent regime of the network activity is characterized by temporal power spectra of the 1/f type. Some properties of the activity time series are compared with well-known results of intermittent regimes in low-dimensional maps.