基于GSA算法改进的K均值聚类  

Improved K-means clustering based on GSA algorithm

作　　者：娄奥 姚敏立 袁丁 LOU Ao;YAO Min-li;YUAN Ding(School of Operational Support,Rocket Force University of Engineering,Xi’an 710025,China)

机构地区：[1]火箭军工程大学作战保障学院,陕西西安710025

出　　处：《计算机工程与设计》2020年第4期1001-1005,共5页Computer Engineering and Design

基　　金：国家自然科学基金项目(61179004、61179005)。

摘　　要：为改善K均值聚类存在的对初始聚心敏感、全局搜索能力弱和凭经验确定聚类数等不足,提出一种基于GSA算法的改进K均值聚类。采用粒子编码策略,把聚类中心集合视作种群粒子,引入GSA搜索聚类质量最好的初始聚类中心,设均方误差为适应度函数,引导全局搜索方向,设置种群成熟度因子避免算法陷入局部最优,引入聚类质量评价指标获取最佳聚类数。通过在4种UCI数据集上做仿真测试,验证了改进后K均值聚类具有较高的正确率和更好的稳定性。To improve the shortcomings of K-means clustering,such as sensitivity to initial concentration,weak global search ability and empirical determination of clustering number,an improved K-means clustering based on gravitational search algorithm(GSA)was proposed.Using particle encoding strategy,cluster centers were regarded as population particles,and GSA was introduced to search the initial cluster centers with the best clustering quality.Mean square error was set as fitness function to guide the global search direction.Population maturity factor was set to avoid the algorithm falling into local optimum.Clustering quality evaluation index was added to obtain the best clustering number.The simulation tests on four UCI datasets show that the improved K-means clustering has higher accuracy and better stability.

关 键 词：K均值 万有引力搜索算法 粒子编码 种群成熟度 最佳聚类数 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]