| Abstract: |
| The single time delayed Hopfield neural network with random connection strengths exhibits dynamic behavior ranging from periodic to chaotic. To understand the underlying mechanisms of rhythm generation, we studied a 10-node Hopfield neural network with one and two time delays with all inhibitory connections. It was found that different connection strengths and time delays produce changes in oscillation frequency and amplitude of the network. In the neural network with a single time delay, we find that oscillation amplitude initially increases exponentially towards saturation and frequency increases linearly with an increase in time delay. Networks with two time delays exhibit a multi-resonance phenomena. Oscillation amplitudes display multiple maxima while the oscillation period shows multiple jumps as the time delay is increased. |
|