EXPLORING ARTIFICIAL INTELLIGENCE AND DATA-DRIVEN TECHNIQUES FOR ANOMALY DETECTION IN CLOUD SECURITY

Abstract

Network security and information system protection from intrusions is a top priority, and intrusion detection systems play a crucial part in this effort. This paper explores an employ of AI and data-driven techniques for enhancing anomaly detection in cloud security, utilizing the CIS-CICIDS2017 dataset. The study develops a DL-CNN model aimed at accurately detecting irregular network traffic patterns and identifying various threats, including DDoS, Heartbleed, and other attacks. Among the models evaluated, the CNN model achieved the best performance, with an accuracy97%, precision93%, recall92%, and an F1-score93%. The CNN outperforms more conventional ML models like RF and GNB, as shown by these data. A finding highlight an effectiveness of AI-driven approaches in strengthening cloud security and provide insights for future research to further improve anomaly detection in dynamic cloud environments.