I.V. Mashkina, A.U. Sentsovа, E.S. Stepanova

The cloud computing technology as a way of granting scalable resources as services to the client via the Internet - is widely adopted now around the world. The architecture of cloud computing system is so difficult that it gains new unknown before vulnerabilities. The transfer to cloud environments leads to emergence of new threats. Application of fuzzy cognitive maps for creation of threats models of unauthorized access and loss information allows to consider all possible threats sources, attacks objects, vulnerabilities. The estimation of predicted information risks in cloud computing system is made on the basis of the developed fuzzy cognitive maps. The results of calculations are used for formation of the sample data set for training the artificial neural network, used for dynamic risk operative estimation.