Automation of the Learning Process of a Neural Network as Classifier in the Formation of a Design Basis of Complex Transport Systems

The method of the automation of the learning process of a neural network as classifier is considered in this paper. This method is based on using the developed algorithm and programs that automatically generate training examples. For this purpose, the methodology of planning a full factorial experiment is used. The classifier is software complex, which includes: a database, a multilayer perceptron, as well as a block of programs providing automatic learning of perceptron. Application neural network as classifier with automated learning allows significantly reduce time spent on the formation of a constructive basis for complex transport systems.