Deep learning, along with neural networks, is an extension of the classical machine learning approach to solving a problem: instead of developing new learners, we can stack together some well-known ones to create an elaborate, but more powerful, learner. This is something similar to the bagging and boosting approach we've seen in the previous section, but with deep learning, this concept is pushed to the limits. Deep learning is nowadays one of the most popular methods of Artificial Intelligence (AI), since it's very effective and general purpose.
The idea of neural networks came from the human central nervous system, where multiple nodes (or, neurons) able to process simple information are connected together to create a network capable to process complex information. In fact, neural networks are named so because they can learn autonomously and adaptively the weights of the model, and they're able to approximate any nonlinear function. In deep learning, the nodes...