VFEFL: Privacy-preserving federated learning against malicious clients via verifiable functional encryption

Key Points

• Privacy-preserving federated learning shows significant resistance to attacks from malicious clients.
• Utilizing cryptographic techniques, the approach ensures data security during distributed learning processes.
• Observational analysis reveals the effectiveness of verifiable functional encryption in safeguarding model integrity.
• This method may enable safer collaborative machine learning without compromising user data privacy.