Privacy-Preserving Epileptic Seizure Detection Using Federated Deep Learning

This paper presents a federated machine learning approach to detect epileptic brain seizures using IoT devices. The proposed system enables distributed training of neural networks across multiple devices while preserving data privacy. Using TensorFlow's Federated Learning framework, a scenario was simulated in which IoT devices locally train models on their data and share updates to refine a global model. Experimental results demonstrate promising accuracy on both training and test datasets, highlighting the potential of federated learning in resource constrained environments.