The first neural networks, which were introduced in the sixties, were inspired by biological neural networks. The idea of a neural network is to map the biological nervous system, that is, how the brain processes information. It consists of layers with interconnected neurons working together. In computer terms, they are also known as an artificial neural network (ANN). With computers, it requires training to make this model learn, the same as a human brain. A neuron in the brain gets activated on receiving a signal from nearby interconnected neurons, and the same applies to an ANN. Recent advances in neural networks has proved that deep neural networks fit very well in pattern recognition tasks, as they are able to automatically extract interesting features and learn the underlying presentation. In this section, we'll refresh ourselves on the fundamental structures and components, from a single perceptron to deep networks.