文章基本信息

标题：Enhancing K-Means Clustering with Bio-Inspired Algorithms
本地全文：下载
作者：Doaa Abdullah ; Hala Abdel-Galil ; Ensaf Hussein 等
期刊名称：International Journal of Computer Science and Network
印刷版ISSN：2277-5420
出版年度：2018
卷号：7
期号：6
页码：361-373
出版社：IJCSN publisher
摘要：Data clustering is considered an important data analysis and data mining technique. It is included in a variety of disciplines such as machine learning, pattern recognition and bioinformatics. K-Means algorithm is a popular clustering algorithm but it suffers from its dependency on its initial centroid locations which fells the algorithm into the local optima. Bio-inspired algorithms are powerful in searching for the global optimal solutions. In this paper, the most recent bio-inspired algorithms; Crow search, Whale optimization, Grasshopper optimization and Salp swarm algorithms are integrated into the K-Means algorithm, to overcome the K-Means drawbacks. The proposed techniques are implemented and applied on eight numerical UCI datasets. Experimental results reveal the capability of the proposed algorithms to find the optimal initial centroid locations which achieve better clustering integrity. Moreover, the results show that the integration of the k-Means with the Crow search algorithm is superior compared to the others bio-inspired algorithms.
关键词：Crow Search Algorithm; Whale optimization Algorithm; Salp Swarm algorithm; Grasshopper Optimization Algorithm; KMeans