Synapses reduction in autoassociative Hopfield network

Abstract

Based on analogy with the oscillator networks, the auto-associative Hopfield network's behavior is investigated for the effect of reducing the connections number. It is shown that the exclusion of connections, with weights modules strictly less than the maximum for a given neuron, significantly improves the network performance. In this case, the allowed share of the distorted input vector elements increases with the network dimension.