I am trying to build a neural network using a group of numbers from a spectrum to determine the existence of a certain peak from the spectrum.
I think that I would like to make this a classification problem: e.g. does this peak exist in the spectrum or not (yes or no). From my spectrum I have (energy, counts), but I am having trouble because my inputs are not -1, 0, 1.
1. Is there a way to use my data without normalizing?
2. Would you recommend using a Hopfield Network?