Continuous attractors of higher-order recurrent neural networks with infinite neurons

This paper investigates continuous attractors of higher-order recurrent neural networks (RNN) with infinite neurons (HRNNwIN). By employing the linearization technique, we present some new criteria on stable, semi-stable, positive semi-global stable and unstable continuous attractors. We also study the continuous attractors of this network under Lognormal distribution except Gaussian distribution. Finally, some simulations are finally carried out to further illustrate the developed theory.