COMPARATIVE EVALUATION OF MACHINE LEARNING MODELS FOR CLOUD INTRUSION DETECTION

Abstract

Cloud computing has become a fundamental technology for delivering scalable and costeffective services over the Internet. However, the shared and dynamic nature of cloud environments makes them highly vulnerable to various cyberattacks, including intrusion attempts, malware injection, and denial-ofservice attacks. Traditional security mechanisms are often insufficient to detect sophisticated and evolving attack patterns. Machine Learning (ML) techniques provide intelligent and adaptive solutions by learning from historical data and identifying anomalous behavior. This paper proposes a Machine Learning-Based Intrusion Detection System (ML-IDS) for cloud security that enhances detection accuracy and reduces false alarms. The proposed system integrates feature selection, supervised learning algorithms, and continuous model training to identify malicious activities effectively. Experimental results demonstrate improved detection performance compared to conventional intrusion detection approaches, making the system suitable for modern cloud environments