Public verifiable privacy protection support vector machine classifier system

The invention provides a public verifiable privacy protection support vector machine classifier system, which is based on a public verifiable homomorphic secret sharing technology, and is characterized in that the public verifiability of a classification result is provided while the privacy of user input and model parameters is protected. The system provides a scheme 1 and a scheme 2, and each of the scheme 1 and the scheme 2 comprises a trusted mechanism for generating a secret key, a user for providing data and obtaining a result, a model provider for providing specific parameters of a support vector machine model, a cloud server which does not need to communicate with each other, and a verifier for verifying the result by any identity. According to the scheme 1, a user needs to carry out interactive communication calculation on a cloud server to obtain a final result, according to the non-interactive scheme 2, the final result can be returned only by transmitting input data on the user, but the cloud serve