Time Encoding Machines are devices that transform a signal from the frequency to the time domain. They're subjects of recent interest due to the hypotheses that neural networks encode information on the precise timing of spikes rather than the amplitude or spike number that was expected. This allows neurons to process high amounts of information on the milliseconds range like vision or sound.
I'm trying to make a neural network using Izhikevich but I still have a long way to go. Current attempts at computing an AND gate using Izhikevich.
|