In this series I'm gonna give you a short overview about the history of artificial neural networks.

1940 - The beginning of neural networks

The beginning of neuroinformatics is often seen in the publication of McCulloch and Pitts. In their work of 1943, they describe how neurons in the brain might be able to work. In 1949 Donald Hebb published his work "The Organization of Behavior ". In this work he described a fundamental concept, how neural networks learn. This was later known as the Hebbsche learning rule and is one of the simplest Learning rules for artificial neural networks. 1951 Marvin Minsky built the first Neurocomputer[1].

1950 - 1960 - The first golden age of neural networks

Between 1957 and 1958, the first successful neurocomputer (called Mark I) was developed by
Frank Rosenblatt and his colleagues. He is considered the inventor of the Perzeptron model.
A short time later, Bernard Widrow developed a new concept in collaboration with his students.
This other type of neural network was called ADALINE. This neural network was trained using the "Stochastic gradient descent" algorithm, which is still in use today. During this period of progress, full-bodied promises were made and neural networks were predicted a glorious future. In 1969 Minsky and Papert's book "Perceptrons" was published, in which they demonstrated that neural networks have significant disadvantages[1].

1970 - The silent development

Despite the publication of Minsky and Papert, the research on neural networks continued.
Especially in the areas of signal processing, pattern recognition and biology the research was pushed further. Many of today's leading experts published their work during this time, including Fukushima and Grossberg. [1]

1980 - New enthusiasm

Between 1983 and 1986, John Hopfield, a world-renowned physicist, began to work on neuronal networks. His two publications in 1982 and 1984 convinced hundreds of researchers worldwide to study neural networks. The Research activity in the field of neural networks increased at an explosive rate after 1986 the books "Parallel Distributed Processing" Volumes 1 and 2 have been published. 1987 was the year when the first conference on neural networks in modern times was held and the "International Neural Network Society" was founded. [2]

Image converted to art by a neural network. source

schematic layout of a artificial neural network

Sources

[1] Yadav, Neha und Yadav, Anupam und Kumar, Manoj. An introduction to neural network methods for differential equations. Springer, 2015.
[2] Kurenkov, Andrey. A ’Brief’ History of Neural Nets and Deep Learning, Part 4.