Intrusion Detection Systems for Securing Cloud Based Financial Services

Abstract

Cloud-based financial services have revolutionized banking and finance, offering scalability, cost-efficiency, and accessibility. However, they are vulnerable to sophisticated cyber threats like DDoS attacks, data breaches, and malware, necessitating robust intrusion detection systems (IDS). This paper explores IDS technologies for cloud environments, focusing on signature-based, anomaly-based, and machine learning (ML) approaches to protect sensitive financial data. Using a simulation-based methodology with the CSE-CICIDS2018 dataset, we propose a hybrid ML-IDS combining Random Forest and deep learning for real-time threat detection. Results show 98.5% accuracy, 97% precision, and 99% recall, outperforming traditional systems by 15%. The study addresses challenges like false positives and scalability, providing a framework for secure cloud financial platforms. Findings emphasize the need for AI-integrated IDS to ensure compliance with regulations like GDPR and RBI guidelines, enhancing trust in digital finance.