CASPIAN JOURNAL
MANAGEMENT AND HIGH TECHNOLOGIES
DETERMINING THE RELEVANCE OF INFORMATION SECURITY THREATS IN INFORMATION SYSTEMS FOR PROCESSING PERSONAL DATA USING THE MATHEMATICAL APPARATUS OF NEURAL NETWORKS
Read | Zhuk Roman V., Dzoban Pavel I., Vlasenko Alexandra V. DETERMINING THE RELEVANCE OF INFORMATION SECURITY THREATS IN INFORMATION SYSTEMS FOR PROCESSING PERSONAL DATA USING THE MATHEMATICAL APPARATUS OF NEURAL NETWORKS // Caspian journal : management and high technologies. — 2020. — №1. — pp. 169-178. |
Zhuk Roman V. - Branch В«Macroregion SouthВ» Ltd Co IC В«SIBINTEKВ», goonerkrd@gmail.com
Dzoban Pavel I. - Kuban State Technological University, antiemoboy@mail.ru
Vlasenko Alexandra V. - Kuban State Technological University, Alex_Vlasenko@list.ru
Describes methods of determining threats to the information security in information systems of personal data processing. Due to the lack of consistency of the existing approved methodology with the one used by the information security threat data Bank (https://bdu.fstec.ru/) the analysis of the parameters of the relevance of threats is carried out and a method for determining the relevance of is threats using the mathematical apparatus of artificial neural networks is proposed. To implement the method, the analysis of artificial neural networks topologies and methods for calculating errors in artificial neural networks is performed. The artificial neural network was developed based on the topology of a multi - layer perceptron with reverse error propagation. Training of the developed information systems was conducted by preparing a training sample. The performance of the developed information systems was compared with the performance of the involved group of experts acting on the existing approved methodology for determining the relevance of is threats in the personal data processing.
Key words: показатель исходной защищенности, потенциал нарушителя информационной безопасности, искусственная нейронная сеть, перцептрон, искусственный нейрон, слой, входной сигнал, выходной сигнал, функция активации, сигмоидальная функция, обучающая выборка, угроза