SecureCloud Guardian: Machine Learning-Driven Privilege Escalation Detection and Mitigation for Cloud Environments

EasyChair Preprint 13083

Abstract

This project employs advanced machine learning to fortify cloud security, specifically targeting and mitigating privilege escalation attacks for a more robust defense mechanism. As cloud adoption rises, so does the risk of privilege escalation attacks. This project addresses vulnerabilities in employee access privileges within cloud services to enhance overall security. Leveraging machine learning, the project enables real-time detection and mitigation of privilege escalation attacks. Techniques like LightGBM, Random Forest, Adaboost, and Xgboost contribute to a dynamic defense against evolving threats. Users and businesses experience heightened data security, fostering trust in cloud computing. Cloud service providers and enterprises gain confidence in a secure online environment, benefiting from the project's security enhancements. And included, a Voting Classifier, amalgamating predictions from Decision Tree, Random Forest, and Support Vector Machine through a "soft" voting approach, enhances the system's performance in detecting and mitigating privilege escalation attacks. Additionally, a user-friendly Flask framework with SQLite integration optimizes user testing, providing secure signup and signin functionalities for practical implementation and assessment

Keyphrases: AdaBoost, Insider Attack, LightGBM, Privilege Escalation, Random Forest, XGBoost, machine learning

@booklet{EasyChair:13083,
  author    = {Shumadhar Reddy Seelam and Sai Vara Prasad Reddy Pulagurla and Naga Aravind Kundeti and M Shobana},
  title     = {SecureCloud Guardian: Machine Learning-Driven Privilege Escalation Detection and Mitigation for Cloud Environments},
  howpublished = {EasyChair Preprint 13083},
  year      = {EasyChair, 2024}}