An Improved Differential Privacy K-Means Algorithm Based on MapReduce
Shunyuan Yao · 2018
In order to solve the low clustering accuracy and the local optimum problem of the traditional differential privacy k-means algorithm, this paper proposed an improved differential privacy K-means algorithm based on MapReduce. The proposed algorithm uses Canopy to select the initial center point, and uses Laplace mechanism to realize the differential privacy protection. The simulation shows that the clustering results of the proposed algorithm outperform the traditional DP K-means in usability and convergence speed.