期刊文献+

基于密度等高线的原型初始化方法 被引量:1

Prototype Initialization Method Based on Density Contour
下载PDF
导出
摘要 对于FCM类算法而言,种子的初始化是保证算法不陷入局部最优值和减少计算时间的关键.本文中提出根据密度分布特性的密度等高线来确定种子的选取不仅可以确定种子的个数,同时可以加快FCM算法的执行. For FCM kinds of algorithms, initialization of seed is a key factor to reduce the computation time and to guarantee avoid ance of the local minimum. A new method of seed selection by density contour is presented in this paper, which can not only determine the numbers of seed but also expedite the execute speed of FCM algorithm.
出处 《南华大学学报(自然科学版)》 2007年第1期38-41,45,共5页 Journal of University of South China:Science and Technology
关键词 密度等高线 聚类 种子初始化 FCM density contour clustering initialization of seed FCM
  • 相关文献

参考文献6

  • 1陈宇.数据挖掘中聚类算法的研究[D].北京:中国科学院数学与系统科学研究院,2001.
  • 2Jiawei Han,Micheline Kamber.Data Mining:Concept and Techniques[M].Copyright by Morgan Kaufmann Publishers,Inc,2001.
  • 3Chaudhuri D,Murthy C A,Chaudhuri B B.Finding a subset of representative points in a data set[J].IEEE Trans.SMC,1992,24(9):1416-1424.
  • 4Chaudhuri D,Chaudhuri B B.A Novel Multiseed Nonhierarchical Data Clustering Technique[J].IEEE Trans.SMC,1997,10(27):871-877.
  • 5Este J,Kriegel H P,Sander J,et al.A Density-Based Algorithm for Discovering Clusters in Large Spatial Databases with noise[M].Proc.KDD'96:226-231.
  • 6George Kollios,Dimitrios Gunopulos,Nick Koudas,et al.Efficient biased Sampling for Apprioximate Clustering and Outlier Detection in Large Data Sets[J].IEEE Trans.KDE,2003,9(15):1170-1187.

同被引文献16

  • 1张敏,于剑.基于划分的模糊聚类算法[J].软件学报,2004,15(6):858-868. 被引量:176
  • 2普运伟,金炜东,朱明,胡来招.核模糊C均值算法的聚类有效性研究[J].计算机科学,2007,34(2):207-210. 被引量:28
  • 3杜廷娜.图像数学形态学分析与处理[J].重庆工商大学学报(自然科学版),2007,24(4):360-362. 被引量:3
  • 4陈健美,宋顺林,陆虎,宋余庆,朱玉全.改进模糊聚类算法及其在入侵检测中的应用[J].东南大学学报(自然科学版),2007,37(4):589-592. 被引量:12
  • 5Kim D W, Lee K H, Lee D On. Cluster validity index for estimation of the optimal number of fuzzy clusters [ J ]. Pattern Recognition, 2004,37 ( 10 ) :2009-2025.
  • 6Gao Xin-bo. Fuzzy clustreing analysis and application [ M ]. Xi'an: Xi'an University of Electronic Science and Technology Press, 2004.
  • 7Zheng Jin-bin, Ouyang Li-jun. Prototype initialization method based on density contour[ J ]. Journal of University of South China ( Science and Technology) ,2007,21 ( 1 ) :38-41,45.
  • 8Chen Jian-mei, Song Shun-lin, Lu Hu, et al. Improved fuzzy clus- tering algorithm and its application to inlrusion detection[ J]. Jour- nal of Southeast University ( Natural Science Edition), 2007,37 (4) :589-592.
  • 9Hung Wen-liang, Yang Miin-Shen, Chen De-hua. Parameter selec- tion for suppressed fuzzy c-means with an application to MRI seg- mentation [ J ]. Pattern Recognition Letters,2006,27 ( 5 ) :424 -438.
  • 10Cai Hui. Connectivity mathematical morphology theory and its ap- plication [D]. Hangzhou:School of Information Science and Engi- neering, Zhejiang University ,2009.

引证文献1

二级引证文献2

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