Information and analytical system for assessing the compliance of educational content specialties ciber security with modern requirements

• Main Authors: Анатолій Степанович Довбиш, Ігор Володимирович Шелехов, Юлія Олексіївна Хібовська, Олександр Вячеславович Матяш
• Format: Article
• Language: English
• Published: National Aerospace University «Kharkiv Aviation Institute» 2021-02-01
• Series: Радіоелектронні і комп'ютерні системи
• Subjects: інформаційно-екстремальна інтелектуальна технологія; машинне навчання; інформаційний критерій; оптимізація; навчальний контент; моніторинг; кібербезпека
• Online Access: http://nti.khai.edu/ojs/index.php/reks/article/view/1353
• ISSN: 1814-4225; 2663-2012

The urgent task of increasing the functional efficiency of machine learning of the information and analytical system (IAS) for assessing compliance with modern requirements of the bachelor level academic disciplines content of the specialty "Cybersecurity" has been solved. A method of information-extremal machine learning IAS has been developed to adapt the educational content of the graduating department to the requirements of the labor market, which allows, when the system is operating in the monitoring mode, to promptly correct the content of the graduating department educational disciplines. The idea of the method is to maximize the information capacity of the IAS in the process of machine learning, which makes it possible to achieve the maximum total probability of making correct classification decisions in the monitoring mode. Because of information-extremal machine learning in the framework of the geometric approach, decisive rules have been constructed that are invariant to the multidimensionality of the recognition feature space. As a criterion for optimizing machine learning parameters, a modification of the Kullback information measure is used, which is a function of the accuracy characteristics of classification decisions. As optimization parameters, the geometric parameters of hyperspherical containers of recognition classes were considered, which were restored in the process of machine learning in the radial basis of the binary space of Hamming features. Simultaneously, the input-training matrix was transformed into a working binary-training matrix, which changes in the process of machine learning by using admissible transformations to adapt the input mathematical description of the system to the maximum reliability of classification solutions. A categorical model of IAS functioning is proposed, based on which an algorithm for information-extreme machine learning of the system with automatic determination of the recognition base class is developed. According to the results of a specialist survey in the field of cybersecurity, an input-structured training matrix was formed and based on the results of physical modeling, the efficiency of the proposed method of information-extreme machine learning IAS was confirmed.