摘要
针对密度峰值快速搜索聚类(Clustering by fast search and find of density peaks,DPC)算法截断距离dc需手动给出的缺陷,提出了布谷鸟优化的密度峰值快速搜索聚类算法(An Improved Cuckoo Search Optimization-based Density Peak Clustering Algorithm,CS-DPC)。引入余弦相似度原理,将方向与实际距离相结合,更好区分两类簇中间区域数据点的归属度。选择5个人工数据集和3个标准UCI数据集进行了实验仿真。
The cutoff distance dc in the clustering by fast search and find of density peaks（DPC）has to be given manually.For the reason,a cuckoo search optimization-based density peak clustering（CSDPC）algorithm is proposed here.The cosine similarity principle is introduced to combine the direction with the actual distance for distinguishing the attribution of the data points of two kinds of clusters.5 groups artificial data sets and 3 groups of UCI standard data sets are used for simulation.
作者
郑虹
周丽媛
韩旭明
ZHENG Hong;ZHOU Liyuan;HAN Xuming(School of Computer Science ＆ Engineering,Changchun University of Technology,Changchun 130012,China)
出处
《长春工业大学学报》
CAS
2018年第3期253-260,共8页
Journal of Changchun University of Technology
基金
国家自然科学基金资助项目(61472049)
吉林省教育厅"十三五"科学技术项目(JJKH20181048KJ)
关键词
截断距离
聚类中心
密度聚类算法
布谷鸟算法
cutoff distance
cluster centers
Density Peak Clustering （DPC）
cuckoo search optimization
