Particle Swarm K-Means++ Clustering Method Based on Multiple Differential Privacy Protection Mechanism

By performing K-Means++ clustering on the samples, the sample data set can obtain a better initial clustering center, and the generation process of this center is transformed using the method of differential privacy, and the Gaussian kernel is used in the process of updating the clustering center. The function performs privacy allocation and adds noise, and finally uses the particle swarm optimization algorithm to optimize and the experimental results show that the algorithm has achieved good prediction results, and the indicators are equal to the original algorithm, so that the algorithm can protect the user’s privacy while ensuring the accurate clustering center.