蜂群K-means聚类算法改进研究  

Research on Bee Colony K-means Clustering Algorithm

作　　者：洪月华[1] 

机构地区：[1]广西经济管理干部学院计算机系,南宁530007

出　　处：《科技通报》2016年第4期170-173,共4页Bulletin of Science and Technology

基　　金：国家社会科学基金项目(15XTQ010);广西高校科学技术研究项目(KY2015YB351)

摘　　要：针对经典K-means聚类算法过于依赖初始聚类中心和易陷入局部最优的不足,提出一种带有学习能力的人工蜂群算法(ABC)与K-means迭代相结合的聚类算法。该算法通过能动态调节的学习权重因子来平衡人工蜂群算法的全局探测与局部搜索能力,同时结合K-means聚类快速的优点,来提高聚类算法的全局寻优能力,降低初始聚类中心对算法聚类质量的影响。通过仿真试验验证,该算法克服了K-means算法的缺点,具有收敛速度快、稳定性强和聚类精度高的优势,得到良好的聚类效果。K-means is a classic clustering algorithm, but it relied heavily on the initial cluster centroids and tends to obtain local optimum only. In this paper, we propose a novel clustering algorithm with a dynamic learning weight factor, which combine artificial bee colony and K-means. This algorithm is efficient since it uses the artificial bee colony approach to balancing the global explore and local search, improves the ability of global optimization, and reduces the effect of initial cluster centroids on the quality of clustering. The experimental evaluations using synthetic datasets show that bee colony K-means algorithm performs more efficiently than traditional K-means, and it has the features of fast convergence, good stability and accurate result.

关 键 词：人工蜂群算法 K-MEANS聚类算法 学习能力 聚类 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]