Build a first Machine Learning model capable of learning its parameters from data

Apply gradient descent to the watering model and discover the limits of a simple score

Transform a score with the sigmoid, discover a first artificial neuron, and make it learn