Evolution of Artificial Neural Networks

Abstract

The history of artificial neural networks originates from the distant 1943, when Warren McCulloch and Walter Pitts formalized the notion of a neural network in a fundamental article on the logical calculation of ideas and nervous activity. This was the starting point in the history of artificial neural networks. Later, a huge number of different architectures of neural networks, learning methods and optimization algorithms were proposed. There was a time when neural networks were left in oblivion, but after the 1980s they were talked about again when John Hopfield introduced his famous full-mesh network. Today, artificial neural networks are indispensable tools in a huge number of tasks, such as tasks of forecasting and analyzing time series, the task of recognizing images and even emotions, classification tasks, natural language recognition, neural networks are used in industry, defense and medicine; many enumerate areas of their application. All this tells us that artificial neural networks have become part of modern life.