This is a barebones script model to understand each calculation step of a basic neural network.
The goal is to feed foward the input values through a single hidden layer and update its weights to reduce the error with the actual output values.
The inspiration of this script is contained in this link.