INTELLIGENT MODELING AND MITIGATION OF EPIDEMIC-SCALE CYBER ATTACKS USING MACHINE LEARNING

Mr. A SRINIVAS; CHINNA PRAVEEN; GANDRAKOTA SANDEEP; DHARMADI BALAJI; JANIGALA SANDEEP KUMAR

doi:10.62643/ijerst.2025.v21.n4.pp661-667

Authors

Mr. A SRINIVAS Author
CHINNA PRAVEEN Author
GANDRAKOTA SANDEEP Author
DHARMADI BALAJI Author
JANIGALA SANDEEP KUMAR Author

DOI:

https://doi.org/10.62643/ijerst.2025.v21.n4.pp661-667

Keywords:

Epidemic Cyber Threats, Cybersecurity, Malware Propagation, Epidemic Modeling, SIR/SEIR Models, Network Immunization, Malware Containment, Machine Learning, Deep Learning, Graph Neural Networks (GNN), Intrusion Detection Systems (IDS), Cyber Attack Prediction, Network Anomaly Detection, Complex Networks, Vulnerability Analysis, Propagation Dynamics, Cyber Defense Strategies, Quarantine Mechanisms, Worm Spreading Models, Threat Modeling, Predictive Analytics, Cyber Epidemics, IoT Security, Attack Mitigation, Intelligent Cyber Defense

Abstract

Epidemic-style cyber security threats exhibit propagation behavior similar to biological outbreaks, making epidemic modeling a powerful tool for analyzing large-scale malware, worms, and network intrusions. Foundational studies on epidemic processes in complex networks [1], epidemic thresholds [3], and network immunization strategies [4], [5], [20] establish how network structure, connectivity patterns, and critical nodes influence the speed and scale of malware spread. Several works adapt classical epidemiological models—including SIR, SEIR, fuzzy models, and multi-stage variants—to cyberspace, enabling accurate prediction and analysis of malware propagation across heterogeneous networks [7]– [10], [19], [21], [22]. Research on self-propagating malware outbreaks, such as WannaCry, further demonstrates the practicality of epidemiological modeling for real-world cyber incidents [2], [6]. Recent advances integrate machine learning and deep learning techniques to enhance malware detection, propagation prediction, and intrusion detection. Graph Convolutional Networks, Graph Neural Networks, and representation learning show significant improvements in modeling propagation behavior and detecting early-stage infections [11]–[13], [23]. Surveys on malware detection and ML-based intrusion detection systems highlight the growing role of data-driven approaches in securing large-scale networks and IoT infrastructures [14]–[16], [24], [25]. Industry insights and technical reports complement academic research by outlining effective containment, isolation, and rapid response techniques crucial for mitigating epidemic cyber threats [17], [18]. Overall, the referenced literature demonstrates that combining epidemic modeling with modern machine learning approaches creates a robust analytical framework for forecasting cyber-attack spread, identifying critical vulnerabilities, and designing proactive defense strategies against fast-moving cyber epidemics.