An intelligent framework for ransomware detection leveraging behavioral Machine Learning and live threat feeds
Sciences des Structures et de la Matiere · 2025
DOI: 10.70974/mat09225070 ↗Kurz gefasst
Fachartikel in der Zeitschrift Sciences des Structures et de la Matiere, 2025. Verfasst von Puneet Chauhan und 1 weiteren Autor:innen.
Abstract (Originalsprache)
Ransomware is a serious and evolving threat to global cybersecurity, utilizing advanced defense techniques that render traditional signature-based defenses ineffective. This paper proposes a new framework for real-time behavioral analysis and response to ransomware (R2BAR), which integrates machine learning with live threat intelligence feeds to enable proactive detection and automated mitigation. The R2BAR framework uses an ensemble approach, which combines a lightweight gradient boosting (XGBoost) model for efficient initial screening with a Long Short-Term Memory (LSTM) network for deep sequential analysis of API call patterns. Detection accuracy is further enhanced by dynamically correlating system behavior with real-time threat intelligence. Experimental evaluation shows that this framework achieves an F1-score of 98.1% and an area under the ROC curve of 0.998, while maintaining a low mean response time (TTR) of 2.35 seconds. This rapid response effectively breaks encryption before significant data loss occurs. The results confirm that the proposed solution addresses the primary limitations of existing methods by striking a balance between high accuracy, operational speed, and interpretability, creating a robust blueprint for the next generation of autonomous ransomware protection systems.
Zitieren
APA
Chauhan, P., & Teotia, S. (2025). An intelligent framework for ransomware detection leveraging behavioral Machine Learning and live threat feeds. Sciences des Structures et de la Matiere. https://doi.org/10.70974/mat09225070
Vancouver
Chauhan P, Teotia S. An intelligent framework for ransomware detection leveraging behavioral Machine Learning and live threat feeds. Sciences des Structures et de la Matiere. 2025. doi:10.70974/mat09225070
BibTeX
@article{Chauhan2025,
author = {Chauhan, Puneet and Teotia, Shashiraj},
title = {An intelligent framework for ransomware detection leveraging behavioral Machine Learning and live threat feeds},
journal = {Sciences des Structures et de la Matiere},
year = {2025},
doi = {10.70974/mat09225070}
}
EUVIMED