Why neural networks have risen to the top of the avant-garde

Scientists have been developing artificial neural networks for more than 70 years. The first attempt to formalize a neural network dates back to 1943 when two American scientists (Warren McCulloch and Walter Pitts) presented a paper on the logical computation of human ideas and neural activity.

Until recently, however, the speed of neural networks was too slow for them to become widespread, and so such systems were mainly used in developments related to computer vision, while other machine learning algorithms were used in other fields.

The laborious and time-consuming part of developing a neural network is training it. For a neural network to be able to correctly solve its tasks, it is required to try how it works on tens of millions of input data sets. It is with the emergence of various accelerated learning technologies that neural networks have become widespread.