Privilege Escalation Attack Detection and Mitigation in Machine Learning

Project Code :TCMAPY1166

Objective

The primary objective of this project is to develop a comprehensive framework for the detection and mitigation of privilege escalation attacks in machine learning systems. This includes designing algorithms and methodologies capable of identifying anomalous behavior indicative of such attacks, as well as implementing countermeasures to prevent unauthorized access and manipulation of privileged functionalities within ML models.

Abstract

The exponential growth in cyber threats due to the proliferation of smart devices poses significant challenges to cybersecurity, especially in cloud environments. This paper addresses the critical issue of privilege escalation attacks through a machine learning-based approach. We propose a systematic method leveraging ensemble learning to detect and classify insider threats, focusing on anomalous occurrences indicative of privilege escalation. Our study utilizes a customized dataset derived from the CERT dataset and evaluates four machine learning algorithms: Decision Tree, AdaBoost, MLP, and Stacking Classifier. By combining these models, we enhance prediction performance and effectively identify security vulnerabilities associated with privilege escalation. Unlike previous studies, our approach emphasizes precise attack identification, contributing to improved security measures in cloud computing environments.

Keywords: Decision Tree, AdaBoost, MLP, Stacking Classifier

NOTE: Without the concern of our team, please don't submit to the college. This Abstract varies based on student requirements.