Compiling...
In the previous note, we discovered that stacking multiple Perceptrons together allows us to solve complex, non-linear problems. When we take this concept and scale it up to include dozens, hundreds, or thousands of neurons organized into structured layers, we create what is known as a Deep Feed-Forward Network.
A Feed-Forward Neural Network (FNN) is the quintessential Deep Learning model. It is called "feed-forward" because the flow of information is strictly unidirectional: Data comes in, flows forward through the mathematical calculations of the network, and a prediction comes out.
