ML from Scratch Multi Layer Perceptron (1/2)

Note: It is important to break the symmetry while weight initialization or else all the neurons will be activated in the same fashion.