Cipher state prediction verification method and device, equipment and storage medium

Abstract

The invention belongs to the technical field of machine learning, and discloses a cipher state prediction verification method and device, equipment and a storage medium. The method comprises the following steps: acquiring a plurality of cipher state data in a cipher state prediction process; splitting the cipher state prediction process, and constructing a plurality of calculation loops; generating a plurality of zero knowledge evidences in batches based on the plurality of cipher state data and the plurality of calculation loops; aggregating the plurality of zero knowledge evidences to obtain a target evidence set; and sending the target evidence set to a verifier, so that the verifier verifies the target evidence set. Through the above mode, the prediction process of cipher state machine learning is converted, the zero-knowledge evidence is generated according to the cipher state data and the prediction process, a verifier verifies the proof to determine whether the prediction process is correct or not, non-interactive public aggregation verification of the cipher state prediction process is realized, and the method is efficiently adapted to a cipher state machine learning scene.

Description

technical field [0001] The present invention relates to the technical field of machine learning, in particular to a dense state prediction and verification method, device, equipment and storage medium. Background technique [0002] In an untrusted machine learning scenario, since the user and the server are not trustworthy to each other, the user (that is, the data owner) usually owns the sample data, and the computing server usually owns the model. During the dense-state machine learning training process, the user encrypts the sample data and sends it to the server, and the server trains to obtain a dense-state model. In the dense-state machine learning prediction process, in order to reduce resource consumption and calculation costs, the malicious server may not perform model inference after receiving the user's prediction request, and return the specified prediction result to the user arbitrarily. On the other hand, in the prediction The former malicious server may modif...

AI Technical Summary

Problems solved by technology

[0005] The main purpose of the present invention is to provide a dense state prediction verification method, device, equipment and storage medium, aiming to solve the problem that users cannot verify the correctness of prediction results, and the existing prediction process verification scheme cannot be efficiently adapted to dense state machines Technical Issues for Learning Scenarios

Images