Neural network operating peculiarities

Abstract

The article analyzes operating peculiarities of the standard neural network for a multilayer perceptron. The network structure and the principle of inverse error propagation algorithm operation are described. The article deals with the notions of artificial neural networks and deep structured learning. Neural networks are the basis of a new and interesting area – deep structured learning. The work covers basic concepts, as well as some code and math, which can help understand and build simple neural networks.