METHOD FOR DETERMINING EFFECTIVE TYPES OF NEURAL NETWORK MODELS FOR RECOGNITION OF CYBER ATTACKS BASED ON PROCEDURAL RULES

Bapiyev I. Khamzina A. Kassymova A. Khazhgaliyeva G. Ramazanova V. Bagisov Zh.
15 May 2022 Little Lion Scientific

Journal of Theoretical and Applied Information Technology
2022 #100 Issue 9 2906 - 2914 pp.

The article touches upon the ways of developing neural network systems for recognizing cyber attacks on network resources of information systems. It is shown that a promising way of such development is the elaboration of rules for determining the effective types of neural network models. The expediency of determining effective types of neural network models based on the consistent application of three rules is substantiated. The first rule allows you to determine the set of valid models from the set of available types of neural network models. The definition of admissibility is implemented on the basis of comparing the minimum possible training time of the type of neural network model with the maximum admissible training time. The second and third rules, formed on the basis of a multi-criteria approach, allow us to select many effective types of neural network models from the set of valid ones, and then determine the most effective type of them.

Cyber Attacks , Network Resources , Neural Network Models , Procedural Rules

Text of the article Перейти на текст статьи

Zhangir Khan West Kazakhstan Agrarian-Technical University, West-Kazakhstan Region, Uralsk, Kazakhstan
Makhambet Utemisov West Kazakhstan University, West-Kazakhstan Region, Uralsk, Kazakhstan

Zhangir Khan West Kazakhstan Agrarian-Technical University
Makhambet Utemisov West Kazakhstan University

10 лет помогаем публиковать статьи Международный издатель

Книга Публикация научной статьи Волощук 2026 Book Publication of a scientific article 2026