My take on the very awesome example by @harvitronix at https://github.com/harvitronix/neural-network-genetic-algorithm If you want to learn how to implement it on Keras using a CNN, I highly suggest that you check his repo out.
Otherwise, if you want to learn how to do it on Sklearn for a normal feedforward neural net, check out the jupyter notebook uploaded in this repo.
- Also, for some reason, my precision, recall, and f1-score metrics aren't being passed as floats. They are extracted in the form of arrays in a list and then passed through a .mean() function. If anyone can help me find out the problem, that would be much appreciated. Cheers!