Optimization of the distributed K-means clustering algorithm based on set pair analysis

The distributed K-means cluster algorithm which focused on multidimensional data has been widely used. However, the current distributed K-means clustering algorithm uses the Euclidean distance as the similarity degree comparison of multidimensional data, which makes the algorithm divides the data set relatively stiff. Aiming at this problem, we present a distributed k-means clustering algorithm(SPAB-DKMC) based on the method of set pair analysis. The results of experiments on the Hadoop distributed platform show that SPAB-DKMC can reduce the number of iterations and improve the efficiency of the distributed K-means clustering algorithm.