期刊文献+

一种基于k-均值的DBSCAN算法参数动态选择方法 被引量:23

k-means based method for dynamically selecting DBSCAN algorithm parameters
下载PDF
导出
摘要 为解决DBSCAN聚类算法的Eps及MinPts参数选择问题,提出一种领域无关的参数动态选择方法。首先,基于k-均值算法对数据集进行初步聚类,聚类中采用最大最小距离方法确定初始聚类中心。其次,针对k-均值聚类结果,计算统计各聚类中样本间距离的分布情况,选择使得具有最大样本对数的距离值作为对应类的Eps值,并通过Eps获得MinPts值。最后,对DBSCAN算法进行改进,使其可根据当前核心点所属k-均值聚类对应的Eps对其运行值进行自适应调整。将上述思想运用于未知协议条件下的比特流聚类分析,结果表明,在无需用户指定Eps及MinPts的条件下,即可获得满意的聚类结果,提高了算法的适用性和准确率。 This paper puts forward a field-irrelative method for dynamically selecting the Eps and MinPts parameters forDBSCAN algorithm.The dataset is first crudely clustered with k-means algorithm using maximum and minimum distanceinitial-centers choosing method.The distance distribution of samples within each k-means cluster is then calculated andanalyzed,choosing the distance which allows maximum point-pair numbers as Eps.The MinPts parameter is also calculatedaccording to the confirmed Eps.Also it improves DBSCAN algorithm to dynamically adjust Eps according to the kmeanscluster to which the current key point belongs.Appling the above ideas to unknown protocol bitstreams clustering,the experiment results demonstrate that the improved DBSCAN can yield satisfied clustering results without manuallyspecifying the Eps and MinPts parameters.The applicability and accuracy of DBSCAN algorithm are improved.
作者 王兆丰 单甘霖 WANG Zhaofeng;SHAN Ganlin(Electronics and Optics Engineering Department, Ordnance Engineering College, Shijiazhuang 050003, China)
出处 《计算机工程与应用》 CSCD 北大核心 2017年第3期80-86,共7页 Computer Engineering and Applications
关键词 聚类 一种经典的基于密度的聚类算法(DBSCAN) 参数选择 K-均值算法 未知协议 clustering Density-Based Spatial Clustering of Applications with Noise(DBSCAN) algorithm parameter selection k-means algorithm unknown protocol
  • 相关文献

参考文献9

二级参考文献100

  • 1周水庚,周傲英,金文,范晔,钱卫宁.FDBSCAN:一种快速 DBSCAN算法(英文)[J].软件学报,2000,11(6):735-744. 被引量:42
  • 2何中胜,刘宗田,庄燕滨.基于数据分区的并行DBSCAN算法[J].小型微型计算机系统,2006,27(1):114-116. 被引量:16
  • 3李杰,贾瑞玉,张璐璐.一个改进的基于DBSCAN的空间聚类算法研究[J].计算机技术与发展,2007,17(1):114-116. 被引量:13
  • 4冯少荣,肖文俊.基于密度的DBSCAN聚类算法的研究及应用[J].计算机工程与应用,2007,43(20):216-221. 被引量:34
  • 5..http://www.ics.uci.edu/mleam/MLSununary.html,.
  • 6MacQueen J.Some methods for classification and analysis of multivariate observations[C]//LeCam L,Neyman J,eds.Proceedings of the Fifth Berkeley Symposium on Mathematics,Statistics and Probability.Berkeley:University of California Press,1967:281-297.
  • 7Leonard Kaufman,Peter J Rousseenw.Finding groups in data:An introduction to cluster analysis[M].New York:Wiley Press,2005.
  • 8Tan P N,Steinbach M,Kumar V 著,范明,范宏建,等译,数据挖掘导论(Introduction to DataMining).北京:人民邮电出版社,2006.
  • 9Ester M,Kriegel H P,Sander J.A density-based algorithm for discovering clusters in large spatial databases with noise[C]//Simoudis E,Hart JW,Fayyad UM,eds.Proceedings of the 2nd International Conference on Knowledge Discovery and Data Mining.Portland:AAAI Press,1996:226-231.
  • 10Ankerst M,Breunig M M,Kriegcl H P.OPTICS:ordering points to identify the clustering structure[C]//Alex Dells,Christns Faloutscs,Shahram Ghandeharizadeh eds.Proceedings of the ACM SIGMOD'99 lnt Conf on Management of Data.Philadelphia Pennsylvania:ACM Press,1999:49-60.

共引文献187

同被引文献191

引证文献23

二级引证文献106

相关作者

内容加载中请稍等...

相关机构

内容加载中请稍等...

相关主题

内容加载中请稍等...

浏览历史

内容加载中请稍等...
;
使用帮助 返回顶部