The Beginner of Machine Revolution

The Birth of the Artificial Neuron

by Eng. Pedro H. Camargos

In 1943 the Warren McCulloch, neurophysiologist, and Walter Pitts, mathematician, introduced the first artificial neural network concepts. The paper, entitled "A Logical Calculus of Ideas Immanent in Nervous Activity," discretized through a simplified computational model the performance of biological neurons through logical propositions. This model becomes known as an artificial neuron.

The working principle of an artificial neuron is simple. The activation is based on logic gates, taken in the graduation course “Digital Electronics” (see Fig. 1). The artificial neuron activates its output when more than a certain number of its inputs are active. Fig. 2 shows the types of artificial neurons introduced by Warren and Walter.

Fig. 1 Logic gates

Fig. 2 Neural artificial neurons performed simple logical computation [1]

The success of artificial neurons stopped until approximately the 1980s. In the early 80s decade news training techniques were invented. In the last decade of the 20th century, artificial neural networks reborn with a new perspective of growth in the scientific milieu. Some factors contributed to this growth: the tremendous increase in computation power, improved training algorithms, and exceptional products based on artificial neural networks.

REFERENCES

[1] Géron, Aurélien. Hands-on machine learning with Scikit-Learn, Keras, and TensorFlow: Concepts, tools, and techniques to build intelligent systems. O'Reilly Media, 2019.

[2] McCulloch, Warren S., and Walter Pitts. "A logical calculus of the ideas immanent in nervous activity." The bulletin of mathematical biophysics 5.4 (1943): 115-133.