Construction of artificial neural network structure :
- Build of an Artificial Neuron
- Build of an MLP (Multi-layers Perceptron)
You have at your disposal a tkinter interface which allows you to test different values ​​of learning rate and iterations for the artificial neuron as well as for the multilayer perceptron and you can enter different layer depths for the MLP with the number of neurons per layer that you want.
Here we use a XOR Dataset for the training, it's a basic dataset but you can modify the X and y for put your own data but don't forget to be careful to respect the dimensions so asn't to have a problem !
Library : Numpy