Multi-Party Computation
Secure multi-party computation (MPC) lets multiple parties jointly compute a function without revealing their private inputs, with only the result disclosed.
Purpose
It enables privacy-preserving collaborative computation among mutually distrusting parties, such as private set intersection, joint statistics, and machine-learning training.
Mechanism
Each party secret-shares its inputs and exchanges shares; the circuit is evaluated interactively over shares (addition gates are added directly, multiplication gates use Beaver triples), and the output is reconstructed at the end. Security models split into semi-honest and malicious.
Use cases
- Private set intersection (PSI)
- Joint healthcare data analysis
- Privacy protection for federated learning