Bayesian Differential Privacy for High-Dimensional Medical Data Optimization

摘要

With the popularization of the Internet and electronic medical records, the security and privacy protection of medical data are increasingly important. The existing differential privacy protection technology cannot effectively handle the problems of the publication of high-dimensional medical data. Based on the current issues of high-dimensional medical data, we proposed a new method, which is a new Bayesian differential privacy protection method, so as to release the high-dimensional medical data safely. Through the experimental evaluation, we confirmed that this method can not only ensure the privacy of the published high-dimensional medical data, but also improve the data availability of the original dataset. The constructed Bayesian network can effectively solve the problem of excessive noise caused by high-dimensional medical data due to scalability and signal-to-noise ratio.