改进的快速DBSCAN算法

Improved fast DBSCAN algorithm

针对DBSCAN算法时间性能低效的问题,分析快速聚类过程中丢失对象的原因,提出一种新的改进算法IF-DBSCAN。该算法在不丢失对象的基础上,通过选取核心对象邻域中的代表对象来扩展类,从而减少邻域查询次数,提高了算法的时间性能。实验结果表明,IF-DBSCAN算法是正确和高效的。

The time performance of Density-Based Spatial Clustering of Applications with Noise （DBSCAN） is inefficient. Concerning this problem, the authors analyzed the reasons of losing object in the process of fast clustering, and proposed a new Improved Fast DBSCAN （IF-DBSCAN） algorithm. On the basis of not losing data object, this algorithm expanded a category by Selecting representative objects from the neighborhood of core data object, so that it reduced the number of regional inquiries and improved the algorithm＇s time performance. The experimental results show that IF-DBSCAN algorithm is correct and efficient.