EUVIMEDEUVIMED entdecken
Crossref

AI-Driven Cyber Defense: A Hybrid Deep Learning Framework for Real-Time Threat Prediction and Response

Lakshmi Kiran Meesala Lakshmi Kiran Meesala

International Journal of Scientific Research in Science and Technology · 01.03.2023 · 916

DOI: 10.32628/ijsrset235845 ↗

Kurz gefasst

Fachartikel in der Zeitschrift International Journal of Scientific Research in Science and Technology, online erschienen am 26.03.2023. Verfasst von Lakshmi Kiran Meesala Lakshmi Kiran Meesala. Die Arbeit wurde bisher 2-mal zitiert (Crossref) und nennt 19 Literaturangaben.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Traditional signature-based and rule-driven intrusion detection systems exhibit critical latency and low recall against zero-day and polymorphic cyber threats. This paper presents CyberHybrid-AI, a novel ensemble architecture integrating Bidirectional Long Short-Term Memory (BiLSTM) networks with Graph Attention Networks (GAT) and a Natural Language Processing (NLP)-driven threat-intelligence module. The proposed system performs continuous behavioral anomaly detection on network flow data while simultaneously extracting semantic threat indicators from unstructured security advisories and dark-web telemetry. The model was trained and validated on the NSL-KDD, CICIDS-2018, and UNSW-NB15 benchmark datasets. CyberHybrid-AI achieved a detection accuracy of 98.7%, F1-score of 0.987, and a mean response latency of 4.2 ms, outperforming state-of-the-art baselines including CNN-LSTM hybrids and Random Forest ensembles by 3.1–7.4 percentage points. Explainability is embedded via SHAP-based attribution, providing transparent, auditable threat scores for SOC analysts. The framework bridges the gap between reactive rule-based defenses and proactive, intelligence-driven cyber resilience. Results confirm statistical significance (p < 0.001) across all benchmark comparisons.

Zitieren

APA

Meesala, L. K. M. L. K. (2023). AI-Driven Cyber Defense: A Hybrid Deep Learning Framework for Real-Time Threat Prediction and Response. International Journal of Scientific Research in Science and Technology, 916. https://doi.org/10.32628/ijsrset235845

Vancouver

Meesala LKMLK. AI-Driven Cyber Defense: A Hybrid Deep Learning Framework for Real-Time Threat Prediction and Response. International Journal of Scientific Research in Science and Technology. 2023:916. doi:10.32628/ijsrset235845

BibTeX
@article{Meesala2023,
  author = {Meesala, Lakshmi Kiran Meesala Lakshmi Kiran},
  title = {AI-Driven Cyber Defense: A Hybrid Deep Learning Framework for Real-Time Threat Prediction and Response},
  journal = {International Journal of Scientific Research in Science and Technology},
  year = {2023},
  pages = {916},
  doi = {10.32628/ijsrset235845}
}
Kostenlos starten →