Formation of neural networks for solving the problem of linear classification

Abstract

There are a large number of methods for solving the linear classification problem, but one of the classic methods for solving this problem is the use of multilayer neural networks and knowledge bases. Therefore, the availability of a convenient and intuitive environment for creating knowledge bases, as well as the creation and training of multilayer neural networks for the analyst is of great importance. In this paper, we consider a model of a multilayer neural network. The analysis of the state of the problem of formation of neural networks for solving the problem of classification is carried out. The main methods and approaches to the formation of multilayer neural networks for solving the problem of classification used for research purposes are considered. The developed model formed the basis of a system that allows you to form and train models of multilayer neural networks for solving problems of linear classification. All this together will help facilitate the work of researchers using neural networks to determine the class of a particular object.