【摘要】 由于地理信息系统中数据的多样性和空间复杂性,在GIS数据空间挖掘时原有的挖掘算法会遇到一些新问题。本文针对城市道路网中交通事故黑点发掘问题,对基于密度的DBSCAN算法进行改进,使改进后的算法能够确保聚类仅在连通空间内进行。最后以某市交通事故数据为例对该方法进行验证,实验结果表明此方法是有效的。更多还原

【Abstract】 Because of the variety of the data and the complexity of space in geographic information system, there are some new prob- lems appearing in the present mining algorithms. In this paper targeting on the digging in traffic accident spot of road net, we im- proved Density- Based DBSCAN algorithm to make sure that process of clustering can be made in connected space. Finally, an experi- ment is conducted basing on traffic accidents data of a city and the experimental result showing our method is practical and effective.更多还原