Social Risk Assessment Mechanism Based on the Neural Networks

This paper is devoted to the analysis of the social risk during the operation and information processing in the complex potentially hazardous objects. There are considered the main factors, that have affect on social risk. In order to perform the risk assesment there was designed and learned the several neural network structures. The method of choosing of the neural network structure which allow obtain the smallest relative error of the result has been proposed and the most acceptable stucture of neural network was chosen. The results of risk assessment shows that the neural network are very close to the results obtained by the dedicated system for risk analysis, so the possibility to use the neural network for the risk analysis in this case is approved.