摘要
基于密度带噪音的空间数据聚类算法,提出了一种改进后的聚类算法。该算法引入了密度分布函数的概念,并采用高斯函数作为影响函数的构成因素。算法以当前具有最大密度的对象作为起点,再从该点的K最邻近结点扩展,直至密度下降到给定的密度阈值时结束。试验测试结果表明:该算法的效果和效率优于传统的基于密度的带噪音的空间数据聚类算法。
An improved clustering algorithm was put forward based on DBSCAN( Density-Based Spatial Clustering of Applications with Noise) that is an traditional and classical clustering algorithm. This algorithm introduced the concept of density distribution function and adopted Gause function as the factor of influence function. Classification was firstly performed from the point that was the maximum density point which extended by KNN( K-Nearest Neighbor algorithm) once again until the density descended to the given density threshold. Experimental results show that the clustering effect and the efficiency of this new algorithm are superior to the traditional DBSCAN.
出处
《河南科技大学学报（自然科学版）》
CAS
北大核心
2014年第5期33-36,109,共4页
Journal of Henan University of Science And Technology:Natural Science
基金
贵州省自然科学基金项目(黔科合丁字[2013]2214号)
贵州省科技厅联合基金项目(黔科合J字LKS[2010]02号
黔科合J字LKS[2009]13号)
关键词
聚类
核心点
密度分布
高斯函数
clustering
core
density distribution
gause function
