Cloud-Based Intrusion Detection System Using Explainable AI

Abstract With the rapid adoption of cloud computing and the exponential growth of network traffic, traditional intrusion detection systems (IDS) face limitations in scalability, adaptability, and transparency. Machine learning–based IDS solutions improve detection accuracy but often behave as black-box models, making their decisions difficult to trust in security-critical environments. This paper proposes a Cloud-Based Intrusion Detection System (CB-IDS) integrated with Explainable Artificial Intelligence (XAI) techniques to provide both high detection performance and interpretability. Using the CICIDS2017 dataset, the proposed system employs Random Forest and XGBoost classifiers to detect malicious network traffic and leverages SHAP (SHapley Additive exPlanations) to explain model predictions at both global and local levels. Experimental results demonstrate high accuracy, precision, recall, and F1-score while offering meaningful explanations of detected intrusions, thereby increasing trust and usability of IDS in cloud environments. Keywords: Intrusion Detection System, Cloud Security, Explainable AI, SHAP, Machine Learning, CICIDS2017