Convolutional Neural Networks (CNNs), or ConvNets, are a special class of feedforward networks; they are primarily used for computer vision tasks, but have also been adapted to other domains with unstructured data, such as natural language processing. As they are feedforward networks, they are very similar to the simple networks that we just learned about; information passes through them in one direction, and they are made up of layers, weights, and biases.
CNNs are the image recognition methods used by Facebook for image tagging, Amazon for product recommendations, and by self-driving cars for recognizing objects in their field of vision. In this chapter, we'll discuss the functions that make CNNs different from standard feedforward networks, and then jump into some examples of how to apply them to a variety of tasks.
In this chapter, we will...
