This survey examines secure multiparty computation techniques for privacy-preserving machine learning in healthcare, suggesting improvements for scalability and regulatory compliance.
Key Points
Secure multiparty computation enhances data privacy in healthcare machine learning, addressing sensitive patient information.
The survey compares secure multiparty computation methods with homomorphic encryption and differential privacy techniques.
Key applications include collaborative model training, privacy-preserving inference, and secure genome analysis in healthcare contexts.
Challenges such as computational overhead and scalability issues are discussed, alongside future research directions.