Modeling and predicting method and device based on cross characteristics, equipment and storage medium

Abstract

The invention discloses a modeling and prediction method and device based on cross characteristics, equipment and a storage medium. The method comprises the following steps of: performing encryption exchange on respective statistics of the first device and the second device according to a preset exchangeable encryption algorithm, calculating to obtain encrypted cross feature statistics corresponding to the first device, and then determining a gradient value corresponding to a model parameter in the first device by using the encrypted cross feature statistics corresponding to the first device,updating the model parameter of the first device based on the gradient value, and performing loop iteration until a preset stop condition is detected to be satisfied to obtain a trained first recommendation model of the first device. Under the condition that the data is not local, the feature intersection between the feature components in the equipment and the feature intersection between the feature components between the equipment are completed at the same time, the complexity of the recommendation model is increased, and therefore the recommendation prediction capacity of the model is improved.

Description

technical field [0001] The present invention relates to the technical field of machine learning, in particular to a modeling and forecasting method, device, equipment and storage medium based on cross-features. Background technique [0002] With the development of artificial intelligence, in order to solve the problem of data islands, people put forward the concept of "federated learning", so that both sides of the federation can also conduct model training to obtain model parameters without giving their own data, and can avoid data The issue of privacy breaches. [0003] Vertical federated learning is to take out the part of users and data with the same participant users but different user data characteristics to jointly train the machine learning model when the data characteristics of the participants are small and the users overlap a lot. For example, two participants, A and B, can use longitudinal federated learning to help A and B build a joint machine learning predict...

AI Technical Summary

Problems solved by technology

However, the existing longitudinal federated learning modeling only considers the respective characteristics of A and B, and the model's recommendation prediction ability is limited

Images