The rapid expansion of distributed computing and cloud storage has increased the demand for intelligent and explainable data security frameworks. Traditional anomaly detection models often function as black boxes, providing little to no insight into decision-making processes, thus limiting their adoption in critical infrastructures where transparency and compliance are mandatory. This study introduces the EnsembleXAI Framework, an integrated architecture combining ensemble learning, explainable artificial intelligence (XAI), federated learning, and distributed edge–cloud computing. The methodology incorporates parallel CNN, LSTM, and Transformer pipelines to extract deep spatiotemporal features, fused through attention gating and ensemble classifiers such as Random Forests and deep neural networks. Interpretability is ensured using SHAP, LIME, and Integrated Gradients, allowing local and global explanations for anomaly detection outcomes. Privacy preservation is achieved via federated learning with differential privacy and homomorphic encryption, while blockchain mechanisms guarantee model update integrity. The proposed framework achieves 98.1% detection accuracy, 97.6% precision, 97.0% recall, and 97.3% F1-score, significantly outperforming baseline CNN, LSTM, and Transformer-only models. Furthermore, explainability fidelity exceeds 93%, ensuring regulatory compliance and user trust. Edge–fog deployment reduces latency to 41.2 ms, enabling real-time inference and dynamic access control for secure cloud storage. Overall, the framework delivers scalable, secure, and interpretable solutions for distributed IoT and cloud ecosystems, advancing the state of trustworthy AI in cybersecurity.
Kamalakkannan et al. (Fri,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: