Changing our model is as easy as redefining our previous build_network() function. Our input layer will stay the same because our input hasn't changed. Likewise, the output layer should remain the same. 

I'm going to add parameters to our network by adding additional hidden layers. I hope that by adding these hidden layers, our network can learn more complicated relationships between the input and output. I am going to start by adding four additional hidden layers; the first three will have 32 neurons and the fourth will have 16. Here's what it will look like:

And here's the associated code for building the model in Keras:

def build_network(input_features=None):
    inputs = Input(shape=(input_features,), name="input")
    x = Dense(32, activation='relu', name="hidden1")(inputs)
    x = Dense(32, activation='relu', name="hidden2")(x)
    x = Dense(32, activation='relu', name="hidden3")(x)
    x = Dense(32, activation='relu', name="hidden4")(x)
    x = Dense(16, activation='relu', name="hidden5")(x)
    prediction = Dense(1, activation='linear', name="final")(x)
    model = Model(inputs=inputs, outputs=prediction)
    model.compile(optimizer='adam', loss='mean_absolute_error')
    return model

As promised, very little of our code has changed. I've bolded the additional lines. The rest of our code can stay the same; however, you often have to train longer (for more epochs) as network complexity increases.