Enhancing Cloud Security: A Hybrid AI Approach for Intrusion Detection Using Convolutional Neural Networks and Stochastic Gradient Descent Algorithms
Malyala Gayatri, Vijayalakshmi Chintamaneni, Revuri Swapna, Malladi Chanti, Lavanya Devarasetty, A. Athiraja · 2024
With the rapid expansion of cloud computing, ensuring the security of data and infrastructure has become paramount. One of the most important functions of Intrusion Detection Systems (IDS) is to protect cloud environments against online attacks. This research study proposes a hybrid Artificial Intelligence (AI) strategy by combining the techniques of Stochastic Gradient Descent (SGD) and Convolutional Neural Networks (CNN) for intrusion detection in cloud computing environment. The proposed methodology integrates the robust feature extraction capabilities of CNN with the optimization power of SGD for efficient model training. By utilizing a comprehensive dataset comprising network traffic patterns and attack scenarios, the hybrid model is trained to accurately classify normal and malicious activities. Through extensive experimentation and evaluation, the proposed approach demonstrates superior performance in detecting intrusions with high accuracy and low false positive rates. Furthermore, the scalability and effectiveness of the proposed method is validated in real-world cloud environments. The research findings highlight the efficacy of hybrid AI techniques in enhancing intrusion detection capabilities, thereby enhancing cloud security and mitigating potential risks in terms of data integrity and confidentiality.