Teacher consensus aggregation learning method based on random response differential privacy technology

Abstract

The invention discloses a teacher consensus aggregation learning method based on a random response differential privacy technology. The method comprises the steps: inputting a query into a teacher model, and obtaining an output tag corresponding to the teacher model, wherein the number of the teacher models is multiple, and the teacher models are different from one another; randomly disturbing the output label to obtain a disturbed label; determining an estimation label corresponding to the query according to the disturbance label, wherein the number of the estimation labels is at least two; and training a student model based on a data set formed by the query and the estimation label and a data set generated by the generative adversarial network to obtain a trained student model. When the student model is migrated, the sensitive data in the sensitive data set cannot be obtained by inquiring and estimating the label, so that the privacy in migration learning is improved.

Description

technical field [0001] The invention relates to the technical field of risk prediction, in particular to a teacher consensus aggregation learning method based on random response differential privacy technology. Background technique [0002] The constructive combination of data-driven learning models and the development of computational and analytical methods has harnessed the full power of massive medical and health data, thereby bringing new insights into healthcare, clinical decision support, and disease risk prediction. new insights. Usually, these algorithms rely heavily on large amounts of well-labeled medical data to build classifiers or predictive models for efficient secondary use, so the quality and quantity of training data obviously have a great impact on the training results. However, due to the low incidence of certain diseases or the long period of medical observation, it is not always possible to use qualified data sets for these diseases. At the same time, t...

AI Technical Summary

Problems solved by technology

[0005] The technical problem to be solved by the present invention is to provide a teacher consensus aggregation learning method based on random response differential privacy technology, aiming at solving the problem of low privacy of transfer learning in the prior art.

Images