新的模糊聚类有效性指标

doi:10.11772/j.issn.1001-9081.2014.08.2166

计算机应用 ›› 2014, Vol. 34 ›› Issue (8): 2166-2169.DOI: 10.11772/j.issn.1001-9081.2014.08.2166

• 第五届中国数据挖掘会议(CCDM 2014)论文 • 上一篇下一篇

新的模糊聚类有效性指标

郑宏亮,徐本强,赵晓慧,邹丽

辽宁师范大学计算机与信息技术学院，辽宁大连116081

收稿日期:2014-05-08 修回日期:2014-05-20 出版日期:2014-08-01 发布日期:2014-08-10
通讯作者: 郑宏亮
作者简介:郑宏亮（1970-），男，辽宁铁岭人，讲师，硕士，主要研究方向：人工智能、数据挖掘；徐本强（1978-），男，黑龙江双城人，讲师，硕士，主要研究方向：人工智能；赵晓慧（1987-），女，辽宁大连人，硕士研究生，主要研究方向：数据挖掘；邹丽（1971-），女，辽宁大连人，副教授，博士，CCF会员，主要研究方向：智能信息处理。
基金资助:
国家自然科学基金资助项目

Novel validity index for fuzzy clustering

ZHENG Hongliang,XU Benqiang,ZHAO Xiaohui,ZOU Li

School of Computer and Information Technology, Liaoning Normal University, Dalian Liaoning 116081, China

Received:2014-05-08 Revised:2014-05-20 Online:2014-08-01 Published:2014-08-10
Contact: ZHENG Hongliang

摘要/Abstract

摘要：

在经典的模糊C均值（FCM）算法中，聚类数需要预先给出，否则算法无法工作，这在一定程度上限制了FCM算法的应用范围。针对FCM算法中聚类数需要预先设定问题，提出了一种新的模糊聚类有效性指标。首先，通过运行FCM算法得到隶属度矩阵；然后，通过隶属度矩阵计算类内紧密性和类间重叠性；最后，利用类内的紧密性和类间的重叠性定义了一个新的聚类有效性指标。该指标克服了FCM算法中类数需要预先设定的缺点，利用该指标可以发现最符合数据自然分布的类的数目。通过对人工数据集和实际数据集的测试表明，对于模糊因子取1.8，2.0和2.2三个不同的常用值，均能发现最优聚类数。

Abstract:

It is necessary to pre-define a cluster number in classical Fuzzy C-means (FCM) algorithm. Otherwise, FCM algorithm can not work normally, which limits the applications of this algorithm. Aiming at the problem of pre-assigning cluster number for FCM algorithm, a new fuzzy cluster validity index was presented. Firstly, the membership matrix was got by running the FCM algorithm. Secondly, the intra class compactness and the inter class overlap were computed by the membership matrix. Finally, a new cluster validity index was defined by using the intra class compactness and the inter class overlap. The proposal overcomes the shortcomings of FCM that the cluster number must be pre-assigned. The optimal cluster number can be effectively found by the proposed index. The experimental results on artificial and real data sets show the validity of the proposed index. It also can be seen that the optimal cluster number are obtained for three different fuzzy factor values of 1.8, 2.0 and 2.2 which are general used in FCM algorithm.

中图分类号:

TP18
TP391

郑宏亮徐本强赵晓慧邹丽. 新的模糊聚类有效性指标[J]. 计算机应用, 2014, 34(8): 2166-2169.

ZHENG Hongliang XU Benqiang ZHAO Xiaohui ZOU Li. Novel validity index for fuzzy clustering[J]. Journal of Computer Applications, 2014, 34(8): 2166-2169.

参考文献

［1］REZAEE B. A cluster validity index for fuzzy clustering ［J］. Fuzzy Sets and Systems, 2010, 161(23): 3014-3025.
［2］ZHANG Y J, WANG W N, ZHANG X, et al. A cluster validity index for fuzzy clustering ［J］. Information Sciences, 2008, 178(4): 1205-1218.
［3］ZALIK K R. Cluster validity index for estimation of fuzzy clusters of different sizes and densities ［J］. Pattern Recognition, 2010, 43(10): 3374-3390.
［4］KIMA 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.
［5］WANG W, ZHANG Y. On fuzzy cluster validity indices ［J］. Fuzzy Sets and Systems, 2007, 158(19): 2095-2117.
［6］YUE S, WANG J, WU T, et al. A new separation measure for improving the effectiveness of validity indices ［J］. Information Sciences, 2010, 180(5): 748-764.
［7］HUANG K. Applications of an enhanced cluster validity index meth-od based on the fuzzy C-means and rough set theories to partition and classification ［J］. Expert Systems with Applications, 2010, 37(12): 8757-8769.
［8］MASSON M, WU T. ECM: an evidential version of the fuzzy C-means algorithm ［J］. Pattern Recognition, 2008, 41(4): 1384-1397.
［9］WU K, YANG M. Robust cluster validity indexes ［J］. Pattern Recognition, 2009, 42(11): 2541-2550.
［10］SUN X, ZHAO Y, WANG H-L, et al. Sensitivity of digital soil maps based on FCM to the fuzzy exponent and the number of clusters ［J］. Geoderma, 2012, 171/172: 24-34.
［11］MITRA S, PEDRYCZ W, BARMAN B. Shadowed C-means: integrating fuzzy and rough clustering ［J］. Pattern Recognition, 2010, 43(4): 1282-1291.
［12］DUNN J C. Well-separated clusters and the optimal fuzzy partitions ［J］. Journal of Cybernetics, 1974, 4(1): 95-104.
［13］BEZDEK J C. Pattern recognition with fuzzy objective function algorithms ［M］. Norwell: Kluwer Academic Publishers, 1981.
［14］BEZDEK J C. Cluster validity with fuzzy sets ［J］. Journal of Cybernetics, 1974, 3(3): 58-73.
［15］BEZDEK J C. Numerical taxonomy with fuzzy sets ［J］. Journal of Mathematical Biology, 1974, 1(1):57-71.
［16］DAVE R N. Validating fuzzy partitions obtained through C-shells clustering ［J］. Pattern Recognition Letters, 1996, 17(6): 613-623.
［17］WINDHAM M P. Cluster validity for the fuzzy C-means clustering algorithm ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1982, 4(4): 357-363.
［18］FUKUYAMA Y, SUGENO M. A new method of choosing the number of clusters for the fuzzy C-means method ［C］// Proceedings of the 5th Fuzzy Systems Symposium. Kobe: ［s.n.］, 1989: 247-250.
［19］GUNDERSON R. Applications of fuzzy ISODATA algorithms to star-tracker printing systems ［C］// Proceedings of the 7th Triennial World IFAC Congress. Helsinki: ［s.n.］, 1978: 1319-1323.
［20］DAVIES D L, BOULDIN D W. A cluster separation measure ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1979, 1(2): 224-227.
［21］GATH I, GEVA A B. Fuzzy clustering for the estimation of the parameters of the components of mixtures of normal distributions ［J］. Pattern Recognition Letters, 1989, 9(2): 77-86.
［22］XIE X, BENI G. A validity measure for fuzzy clustering ［J］. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1991, 13(4): 841-847.
［23］WINKLER R, KLAWONN F, KRUSE R. Fuzzy C-means in high dimensional spaces ［J］. International Journal of Fuzzy System Applications, 2011, 1(1): 1-16.
[24]ZAHID N, LIMOURI M, ESSAID A. A new cluster validity for fuzzy clustering [J]. Pattern Recognition, 1999, 32(7): 1089-1097.
[25]WU K, YANG M. A cluster validity index for fuzzy clustering [J]. Pattern Recognition Letters, 2005, 26(9): 1275-1291.
[26]PAKHIRA M K, BANDYOPADHYAY S, MAULIK U. Validity index for crisp and fuzzy clusters [J]. Pattern Recognition, 2004, 37(3): 487-501.

[1]	杜航原郝思聪王文剑. 结合图自编码器与聚类的半监督表示学习方法[J]. 计算机应用, 0, (): 0-0.
[2]	陈露张晓霞于洪. 基于先验知识的非负矩阵半可解释三因子分解算法[J]. 计算机应用, 0, (): 0-0.
[3]	韩舒宁徐敏董学士林青沈凡凡. 混合伊藤算法求解多尺度着色旅行商问题[J]. 计算机应用, 0, (): 0-0.
[4]	李晓杰崔超然宋广乐苏雅茜吴天泽张春云. 基于时序超图卷积神经网络的股票趋势预测方法[J]. 计算机应用, 0, (): 0-0.
[5]	张建严珂马祥. 基于神经网络的复杂垃圾信息过滤算法分析[J]. 计算机应用, 0, (): 0-0.
[6]	邱云志汪廷华戴小路. 双重特征加权模糊支持向量机[J]. 计算机应用, 0, (): 0-0.
[7]	李宗正周恺卿丁雷欧云. 基于基因交换的自适应人工鱼群算法[J]. 计算机应用, 0, (): 0-0.
[8]	刘清华廖士中. 基于随机素描方法的在线核回归[J]. 计算机应用, 0, (): 0-0.
[9]	张小清王晨曦吕彦林耀进. 基于ReliefF的层次分类在线流特征选择算法[J]. 计算机应用, 0, (): 0-0.
[10]	于婉莹梁美玉王笑笑陈徵曹晓雯. 基于深度注意力网络的课堂教学视频中学生表情识别与智能教学评估[J]. 计算机应用, 0, (): 0-0.
[11]	黄勇康梁美玉王笑笑陈徵曹晓雯. 基于深度时空残差卷积神经网络的课堂教学视频中多人课堂行为识别[J]. 计算机应用, 0, (): 0-0.
[12]	康猛蒙祖强. 基于局部条件区分能力的高效属性约简算法[J]. 计算机应用, 0, (): 0-0.
[13]	谢鑫张贤勇王旋晔唐鹏飞. 变精度邻域等价粒邻域决策树构造算法[J]. 计算机应用, 0, (): 0-0.
[14]	刘忠慧王梓宥闵帆. 近似概念的遗传生成算法及其推荐应用[J]. 计算机应用, 0, (): 0-0.
[15]	潘仁志钱付兰赵姝张燕平. 基于卷积神经网络交互的用户属性偏好建模的推荐模型[J]. 计算机应用, 0, (): 0-0.

新的模糊聚类有效性指标

Novel validity index for fuzzy clustering

PDF

可视化

摘要/Abstract

引用本文

使用本文

参考文献

相关文章 15

编辑推荐

Metrics