Introducing deep neural networks

Nature has always inspired mathematicians and engineers to devise appropriate algorithms, designs, and artifacts for any given problem. So incorporating solutions that have proven themselves over millennia seems like a good idea. We can refer to numerous examples such as bats’ echolocation that inspired human-made sonars, the high-speed trains that have a shape that resembles the elongated beak of kingfisher birds to prevent sonic booms, the flight of drones as a flock of birds to avoid collisions, and many more.

Throughout this chapter, we discussed many times how algorithms learn from data. What is more natural than to think that emulating the human brain and its functionalities can enhance artificial cognition? Exploiting the mode of operation of this astonishingly complex organ of the human nervous system might permit the creation of sophisticated algorithms in any domain. This section provides a gentle introduction to the topic and presents...