Federated learning classification tree construction method, model construction method and terminal equipment

Abstract

The embodiment of the invention relates to a federated learning classification tree construction method, a model construction method and terminal equipment. The method comprises the steps of carryingout the classification of a user sample set according to a first classification feature and a first classification threshold, and obtaining at least two groups of user sample subsets; adding confuseduser samples into the first subset to obtain a second subset; calculating the sum of the first encryption gradient values corresponding to the second subset; after the primary encryption gradient value is subjected to secondary encryption, sending the primary encryption gradient value and the second subset to a second data provider together; receiving the sum of the second encryption gradient values; performing primary decryption on the sum of the second encryption gradient values, and feeding back the sum of the second encryption gradient values, the sum of the first encryption gradient values, the information of the first classification feature and the information of the first classification threshold to the second data provider; receiving information of the optimal classification feature and information of an optimal classification threshold corresponding to the optimal classification feature; and classifying the first subset according to the information of the optimal classification feature and the information of the optimal classification threshold corresponding to the optimal classification feature to form branch nodes of a federated learning classification tree, and constructing the federated learning classification tree.

Description

technical field [0001] The embodiment of the present invention relates to the field of computer technology, and in particular to a federated learning classification tree construction method, a model construction method, and a terminal device. Background technique [0002] With the development of machine learning technology, more and more machine learning methods have been widely used in financial, medical and other fields. In machine learning, the data determines the upper limit of the model effect. In order to further improve the accuracy of the model, federated learning methods that aggregate multi-party data for model training have begun to appear. [0003] Since different companies may have the characteristics of different dimensions of the same group of users, aggregating different characteristics can effectively improve the effect of machine learning. The goal of vertical federated learning is to aggregate feature data of different dimensions from multiple parties, sh...

AI Technical Summary

Problems solved by technology

If the model lacks interpretability, we cannot intuitively judge whether the model has captured meaningful features, nor can we know how well the model generalizes to other sample predictions, which may make the model's prediction results questionable

Images