Neural network architectures

Neural networks are currently one of the most modern creative and interesting areas of knowledge. The use of neural networks allows solving many problems when the use of other approaches is almost impossible due to their complexity. Todaty, there are a large number of neural network architectures available, including simple ones like perceptron and neural networks with trillions of parameters. Such a variety of neural network architectures is determined by the tasks they face. In addition, the scope of neural networks is also growing due to the emergence of the possibility of their application in connection with increasing data volumes and computing power. The article provides an overview of the main of neural networks architectures, which serve as the basis for development of the modern neural networks, that can be used in many aspects of everyday life.