An Adaptively Disperse Centroids K-Means Algorithm Based on MapReduce Model
Bin Wang, Zheng Lv, Jinwei Zhao, Xiaofan Wang, Tong Zhang · 2016
K-means is a clustering algorithm which is used widely. Its clustering results heavily depend on the initial centroids. An adaptive method for disperse centroids is proposed to improve the stability and accuracy of the clustering result. The Adaptively Disperse Centroids K-means Algorithm (ADC-K-means) is implemented using MapReduce model on hadoop platform, and it is compared with the k-means of Mahout which is a sub-project of hadoop. The experimental result shows that proposed algorithm is effective.