【摘要】 在DBSCAN算法中需要人工输入Eps和MinPts两个参数,因而聚类过程需要用户的干预才能进行,导致聚类结果的准确度直接取决于用户对参数的选择。鉴于此,本研究提出了一种新的Eps和MinPts参数的确定方法,避免了聚类过程中的人工干预,实现了聚类过程的全自动化。理论分析和实验结果表明,该方法能够选择合理的Eps和MinPts参数并得到较高准确度的聚类结果。更多还原

【Abstract】 In the DBSCAN algorithm,the parameters Eps and MinPts’ input is needed manually.So that the process of clustering requires user’s intervention,whereby leading to the accuracy of clustering results which depends directly on the user’s selection of parameters.For this reason,this paper proposes a new method to determine the parameters Eps and MinPts,so as to avoid the manual intervention in the process of clustering and realize automation in the clustering.Theoretical analysis and experimental results show that the method can choose the rational parameters Eps and MinPts,and get a higher accuracy of clustering results.更多还原