计算机工程与应用 ›› 2007, Vol. 43 ›› Issue (20): 216-221.

• 工程与应用 • 上一篇    下一篇

基于密度的DBSCAN聚类算法的研究及应用

冯少荣1,2,肖文俊1   

  1. 1.华南理工大学 计算机科学与工程学院,广州 510640
    2.厦门大学 信息科学与技术学院,福建 厦门 361005
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2007-07-11 发布日期:2007-07-11
  • 通讯作者: 冯少荣

Research and application of DBSCAN clustering algorithm based on density

FENG Shao-Rong1,2,XIAO Wen-Jun1   

  1. 1.School of Computer Science and Engineering,South China University of Technology,Guangzhou 510641,China
    2.College of Information Science and Technology,Xiamen University,Xiamen,Fujian 361005,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-07-11 Published:2007-07-11
  • Contact: FENG Shao-Rong

摘要: 首先对DBSCAN(Density Based Spatial Clustering of Applications with Noise)聚类算法进行了深入研究,分析了它的特点、存在的问题及改进思想,提出了基于DBSCAN方法的交通事故多发点段的排查方法及其改进思路,并且给出了实例以说明处理过程及可行性。实验结果表明本文提出的方法可以大大提高交通事故黑点排查效率。

关键词: 聚类分析, DBSCAN, 交通事故多发点(段), 数据挖掘

Abstract: This paper first researches DBSCAN clustering algorithm,and analyzes characteristics and existing problems of the DBSCAN algorithm and improved idea.Evaluation method of the traffic accident black spots and an improved thought based on DBSCAN are proposed.In order to illuminate course of processing and feasibility,an example is presented.The experimental result demonstrates that this paper method can greatly enhance the working efficiency of evaluation of the traffic accident black spots.

Key words: clustering analysis, Density Based Spatial Clustering of Applications with Noise(DBSCAN), prone location of traffic, data mining