Corporate Infocommunication Networks Research by Neural Networks Method with Calculation of Generalization
I. O. Yuskov, E. P. Stroganova
Abstract
One of the main options for using the future neural network is to use it in infocommunications in the application for corporate networks. To achieve such goals, it is possible to create algorithms of actions to enhance and improve neural corporate networks. The network generalization, in turn, is described as a logical segment of the physical network. Learning on the basis of some data and the correct application of the acquired knowledge on other data is called generalization. The proposed option offers to increase the efficiency of optimization for the corporate network and various selection procedures for optimal technological conditions of operation. Based on the applied analysis, formulas were calculated and a conclusion was made about the future of this method in this field.