EARLY IDENTIFICATION OF CYBER THREATS BY DETECTING ANOMALIES BASED ON ARTIFICIAL INTELLIGENCE
Keywords:
artificial intelligence, cybersecurity, anomaly detection, cyber threats, machine learning, network traffic, proactive protection, Autoencoder, zero-day attacks.Abstract
This scientific article is devoted to the issue of early identification of cyber threats by detecting anomalies based on artificial intelligence. The complexity and dynamic nature of cyberattacks in modern information and communication systems significantly reduce the effectiveness of traditional signature-based security systems. Therefore, this study examines approaches to anomaly detection based on artificial intelligence and machine learning technologies for identifying cyber threats.In the research process, unsupervised and semi-supervised machine learning methods were applied based on the analysis of normal network traffic behavior. In particular, anomaly detection mechanisms were developed using clustering algorithms and deep learning-based Autoencoder models. Network traffic data were cleaned, normalized, and used to train models primarily on normal activity patterns due to the limited availability of labeled attack data in real-world environments.
The obtained results demonstrate that artificial intelligence-based anomaly detection approaches enable the early identification of unknown and zero-day cyberattacks. These approaches support the transition from reactive to proactive cybersecurity mechanisms and provide a scientific foundation for integrating artificial intelligence-based anomaly detection methods into real network infrastructures.
References
1. Sommer R., Paxson V. Outside the Closed World: On Using Machine Learning for Network Intrusion Detection. IEEE Symposium on Security and Privacy.
2. Chandola V., Banerjee A., Kumar V. Anomaly Detection: A Survey. ACM Computing Surveys.
3. Goodfellow I., Bengio Y., Courville A. Deep Learning. MIT Press.
4. Patcha A., Park J. An overview of anomaly detection techniques: Existing solutions and latest technological trends. Computer Networks.
5. Lakhina A. et al. Mining Anomalies Using Traffic Feature Distributions. ACM SIGCOMM.
