UDC 004.032.26
DOI: 10.36871/2618-9976.2022.11.003

Authors

Sergey V. Zaytsev
Postgraduate Student of the Department of «System Analysis in Economics», Financial University under the Government of the Russian Federation, Moscow, Russia

Abstract

The formation of neural networks has caused a lot of enthusiasm and criticism. Some studies turned out to be optimistic, others – pessimistic. For many tasks, such as image identification, no dominant approaches have yet been made. It is necessary to try to realize the possibilities, prerequisites and scope of using different approaches and maximize their additional advantages for the subsequent formation of intelligent systems. Every year, new types of neural networks, new types of training, and the implementation of samples appear. And every year there are new tasks that humanity needs to solve, for this we create new networks, thereby neural networks begin to grow together with humanity. Despite their disadvantages, they are a very convenient tool that in the future will be able to help humanity reach new horizons.

Keywords

Neural networks, Self-learning algorithms, Applied solutions, Genetic algorithms