This mini-module contains two main Python classes that pertain the domain of fully-connecter neural-networks.
The trainFCNN
class wraps the code to create some synthetic datasets and learn a neural network model with any of the following libraries: Scikit-learn, Keras, TensorFlow and PyTorch.
The visFCNN
class aims at visualizing what happens inside the network with some colourful charts.
The visualizations have been created with pure Matplotlib.