A friendly little challenge
- And another thing, I'd like to see if it's possible to create a neuron that changes its own waveform pattern when the weights at its inputs are altered. From square to sinusoid to cosine to triangular to sawtooth, etc. This neuron would have two basic inputs: enable input and control input. The enable input triggering an active process when activated, and the control determining the waveform pattern the neuron would transmit. Consider this a challenge. I just want to see if it can be done. I want to try some things myself, but I figure it will go faster if everyone else takes a shot at it too. Here are the rules to qualify the neuron:
*The circuit must have a control, or selector, input.
*The circuit must be able to use this to select any type of waveform.
*The circuit must also be able to compare the waveforms at both inputs and dynamically alter its reaction based on such data.
*The circuit must have a feedback output(preferably cosine or irregular waveforms). This would help determine the state of the network as a whole.
You don't have to respond to this post right away, but if you ever fell like it, give it a shot.