Analysis and Research Based on Differential Privacy Protection Related Algorithms

Aiming at the problem that traditional privacy protection methods can not cope with malicious analysis under any background knowledge, a differential privacy protection model is proposed. This article first introduces the concept of differential privacy, and then introduces the relevant definitions and properties. The article mainly introduces differential private algorithm, differential private random forest algorithm and differential private algorithm, and describes the principle and flow of each algorithm in detail. We will test the performance of related algorithms in a cloud environment. In the experiment, we compare and analyze the accuracy of the algorithm by setting different parameter values, and finally give the corresponding conclusions.