Machine learning model training method, prediction method and device based on artificial intelligence

Abstract

The invention provides a machine learning model training method and device based on artificial intelligence, a prediction method and device, electronic equipment and a storage medium. The method comprises the steps of performing fusion processing on private training data of a first participant participating in machine learning model training to obtain local information of the first participant; wherein the private training data of the first participant comprises a private weight parameter held by the first participant; private training data of the first participant and private training data ofa second participant participating in machine learning model training are combined for privacy protection processing, and shared intermediate information is obtained; and determining the gradient ofthe machine learning model corresponding to the first participant according to the shared intermediate information and the local information of the first participant, and updating the private weight parameter corresponding to the first participant along the descending direction of the gradient. According to the invention, the security of data held by the participant can be improved in the model training and prediction process.

Description

technical field [0001] The present invention relates to artificial intelligence technology, in particular to an artificial intelligence-based machine learning model training method, prediction method, device, electronic equipment and storage medium. Background technique [0002] Artificial Intelligence (AI) is a theory, method, technology and application system that uses digital computers or machines controlled by digital computers to simulate, extend and expand human intelligence, perceive the environment, acquire knowledge and use knowledge to obtain the best results. Machine learning (ML, Machine Learning) is an important branch of artificial intelligence. It mainly studies how computers simulate or implement human learning behaviors to acquire new knowledge or skills, and reorganize existing knowledge structures to continuously improve their own performance. . [0003] In the training process of the machine learning model, different participants may be involved, and eac...

AI Technical Summary

Problems solved by technology

In order for all participants to share the model, in the solutions provided by related technologies, the trained model is usually split to each participant according to the attributes owned by each participant. However, if a participant can provide a Data, even if there is only one non-missing value for a certain attribute, then the participant can obtain all the weight parameters of the model, resulting in the possibility of leakage of private data of other participants, and low data security

Images