<ul><li>Federated learning is a machine learning method that supports training models on decentralized devices or servers, without data exchange.</li><li>A new approach, FAS, employs selective encryption, homomorphic encryption, differential privacy, and bit-wise scrambling to minimize data leakage.</li><li>Experiments show FAS is up to 90% faster than fully homomorphic encryption on model weights, saving up to 46% in total execution time.</li><li>FAS achieves good execution performance and similar security results compared to a state-of-the-art federated learning approach.</li></ul>

A Selective Homomorphic Encryption Approach for Faster Privacy-Preserving Federated Learning

Discover more