A digital perceptron that also solves the XOR problem
Already have an account? Log in.
To make the experience fit your profile, pick a username and tell us what interests you.
The next questions I will be addressing in the near future are as follows:
Since the error can also be only 0 or 1, the occurrence of an error means that the inverted input vector must be true. This would then be the backpropagation.
Thanks you. Unfortunately, I can't say offhand, because the digital perceptron has only two inputs at the moment and can learn all logic gates in three iterations. So you would have to build a neural network with it.
Oh I didn't realize that perceptron is 2 to 1 mux of sorts, and it needs to get generalized to a M to N for such problems.
M. Bindhammer
Al Williams
Dana Myers
Tim
This is interesting project and idea... If I had a digital perceptron with 9 inputs and 5 outputs, how many samples would I need to present it to teach is that inputs are A+B+carry and output is Y+carry out?