Research and application of DBSCAN clustering algorithm based on density

摘要

摘要：首先对 D BSCA N（D ensity Based Spatial Clustering of A pplications with N oise）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 D BSCA N 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。A bstract：This paper first researches D BSCA N clustering algorithm，and analyzes characteristics and existing problem s of the D B-SCA N algorithm and im proved idea.Evaluation m ethod of the traffic accident black spots and an im proved thought based on D B-SCA N are proposed.In order to illum inate course of processing and feasibility，an exam ple is presented.The experim ental resultdem onstrates that this paper m ethod can greatly enhance the working efficiency of evaluation of the traffic accident black spots.