EUVIMEDEUVIMED entdecken
Crossref

Conceptual Principles for Developing Specialised Neural Network Modules for Cyberattack Detection Based on Vector Classification and Economic Damage Assessment

Oleksandr Kushnerov, Dmytro Likholietov, Ruslan Andriienko

Economic sustainability and business practices · 25.05.2026 · 3 (2) · 18

DOI: 10.21272/esbp.2026.2-03 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Economic sustainability and business practices, online erschienen am 25.05.2026. Verfasst von Oleksandr Kushnerov und 2 weiteren Autor:innen. Die Arbeit nennt 19 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

This article presents comprehensive research results aimed at developing conceptual foundations for specialised neural network modules to detect cyberattacks. The proposed approach is based on the synergistic integration of a detailed vector-based classification of cyberattacks and methodologies for quantitatively assessing associated economic damages. To achieve this goal, an in-depth theoretical and methodological analysis was conducted, including a critical review and evaluation of the suitability of existing cyberattack taxonomies – such as MITRE ATT&CK, CAPEC, OWASP Top 10, and ENISA Threat Landscape reports – for characterizing attack vectors, velocity, and complexity across various infrastructures (Information Technology, Operational Technology, cloud environments, and the Internet of Things). Concurrently, quantitative (e.g., FAIR, ROSI, VaR) and qualitative methodologies for assessing the economic impact of cyber incidents were analysed, enabling the identification of their advantages and limitations for prioritising cybersecurity defence tasks. The authors’ key contribution lies in developing a novel vector-oriented classification framework that systematises attacks based on key attributes (e.g., entry point, propagation mechanism, target specificity) and integrates quantitatively measurable metrics of attack velocity and complexity into a unified vector. The scientific novelty consists of formulating conceptual foundations, derived from this synthesis, which directly link the refined attack vector to the selection of an optimal neural network architecture – such as convolutional, recurrent, graph neural networks, and transformers – as well as in using economic risk profiles calculated for each vector as the primary criterion for prioritizing the development of the corresponding detection modules. The practical significance of this research is in creating an integrated conceptual model that systematises the entire development lifecycle, from threat analysis and economic risk assessment to the design, testing, and iterative refinement of neural network-based solutions. The research also explores advanced feature engineering strategies and architectural principles to ensure module adaptability and resilience against data imbalance, concept drift, and adversarial attacks. The principal conclusion is that the proposed comprehensive approach establishes a robust scientific and methodological foundation for developing a new generation of cyberattack detection systems that are not only technically advanced but also economically justified and capable of effectively countering the dynamic landscape of contemporary cyber threats.

Zitieren

APA

Kushnerov, O., Likholietov, D., & Andriienko, R. (2026). Conceptual Principles for Developing Specialised Neural Network Modules for Cyberattack Detection Based on Vector Classification and Economic Damage Assessment. Economic sustainability and business practices, 3(2), 18. https://doi.org/10.21272/esbp.2026.2-03

Vancouver

Kushnerov O, Likholietov D, Andriienko R. Conceptual Principles for Developing Specialised Neural Network Modules for Cyberattack Detection Based on Vector Classification and Economic Damage Assessment. Economic sustainability and business practices. 2026;3(2):18. doi:10.21272/esbp.2026.2-03

BibTeX
@article{Kushnerov2026,
  author = {Kushnerov, Oleksandr and Likholietov, Dmytro and Andriienko, Ruslan},
  title = {Conceptual Principles for Developing Specialised Neural Network Modules for Cyberattack Detection Based on Vector Classification and Economic Damage Assessment},
  journal = {Economic sustainability and business practices},
  year = {2026},
  volume = {3},
  number = {2},
  pages = {18},
  doi = {10.21272/esbp.2026.2-03}
}
Kostenlos starten →