TY - GEN
T1 - Securing Cloud Networks Against Data Exfiltration Using Random Forest Techniques
AU - Saroj, Nancy
AU - Singh, Sunil K.
AU - Kumar, Sudhakar
AU - Rajak, Siya
AU - Bawa, Sumit
AU - Arya, Varsha
AU - Chui, Kwok Tai
AU - Gupta, Brij B.
N1 - Publisher Copyright:
© 2025 IEEE.
PY - 2025
Y1 - 2025
N2 - As cloud computing becomes increasingly integral to modern infrastructure, the risk of unauthorized data transfer poses a significant threat to security. Random Forest Regression, which estimates a continuous anomaly risk score to detect behaviors indicative of data exfiltration in cloud platforms. Key features, including attack type, severity level, and session metadata, were extracted and processed to enhance predictive performance. The model achieved a Mean Absolute Error (MAE) of 0.154, a Mean Squared Error (MSE) of 0.23761, and an R2 score of 0.778, underscoring both the complexity of rare event detection and the model's potential for real-time anomaly recognition. This approach highlights the value of adaptive, data-driven techniques in enhancing cloud security systems against evolving threats.
AB - As cloud computing becomes increasingly integral to modern infrastructure, the risk of unauthorized data transfer poses a significant threat to security. Random Forest Regression, which estimates a continuous anomaly risk score to detect behaviors indicative of data exfiltration in cloud platforms. Key features, including attack type, severity level, and session metadata, were extracted and processed to enhance predictive performance. The model achieved a Mean Absolute Error (MAE) of 0.154, a Mean Squared Error (MSE) of 0.23761, and an R2 score of 0.778, underscoring both the complexity of rare event detection and the model's potential for real-time anomaly recognition. This approach highlights the value of adaptive, data-driven techniques in enhancing cloud security systems against evolving threats.
KW - Anomaly Detection
KW - Cloud Security
KW - Cybersecurity
KW - Data Exfiltration
KW - Machine Learning
KW - Predictive Modeling
KW - Random Forest Regression
KW - Real-Time Threat Detection
UR - https://www.scopus.com/pages/publications/105036829917
U2 - 10.1109/ICATEI67676.2025.11405243
DO - 10.1109/ICATEI67676.2025.11405243
M3 - Conference contribution
AN - SCOPUS:105036829917
T3 - ICATEI 2025 - International Conference on Advanced Technologies in Energy and Informatic
SP - 166
EP - 171
BT - ICATEI 2025 - International Conference on Advanced Technologies in Energy and Informatic
T2 - 2025 International Conference on Advanced Technologies in Energy and Informatics, ICATEI 2025
Y2 - 22 October 2025 through 22 October 2025
ER -