Building a model

Now we can move on and build our predictive model. But before jumping into the interesting code, we must know the theory that surrounds a few important things.

Introducting Keras layers

The core building blocks of a neural network model in Keras is its layers. Layers are basically data-processing filters that contort the data they are fed into more useful representations. As we will see, prominent architectures of neural networks mostly vary in the manner in which layers are designed, and the interconnection of neurons among them. The inventor of Keras, Francois Chollet, describes this architecture as performing a progressive distillation on our data. Let's see how this works:

#Simple Feedforward Neural...