Improvement of distributed clustering algorithm based on min-cluster

详细信息    查看全文

• 作者：Hao Zhang^a ; ^b ; ^{hao19825@tlu.edu.cn" class="auth_mail" title="E-mail the corresponding author} ; GuangLong Dai^a ; ^{gldai@aust.edu.cn" class="auth_mail" title="E-mail the corresponding author}
• 关键词：Min-cluster ; K-Dmeans algorithm ; Clustering ; Distributed clustering
• 刊名：Optik - International Journal for Light and Electron Optics
• 出版年：2016
• 出版时间：April 2016
• 年：2016
• 卷：127
• 期：8
• 页码：3878-3881
• 全文大小：581 K

文摘

This article, aiming at the inefficiency clustering result of the raw data, presents a new way, center – edge system frame structure, to improve K-Dmeans algorithm. Also min-cluster, as a new concept, is proposed when the edge node of dataset is clustering, which makes the data of edge node progress the refinement of clustering regularly, and transfers the result to the center node directly. The clustering data will be integrated again by the center node. In the privacy protect distributing clustering arithmetic. We adopt the concept of min-cluster from data string, and join in multi-security protocols. From the experiment, the distributed clustering algorithm is proved to be with high accuracy, with low time complications, and the last but most important security.