智能单粒子优化算法在聚类分析中的应用*

南京大学学报(自然科学版) ›› 2011, Vol. 47 ›› Issue (5): 578–584.

智能单粒子优化算法在聚类分析中的应用*

陈永彬1，张琢1,2**

出版日期:2015-05-04 发布日期:2015-05-04
作者简介: (1.东北师范大学理想信息技术研究院，长春,130024;
2.吉林远程教育重点实验室，长春,130024
基金资助:

An application of intelligent single particle optimizer in cluster analysis

Chen Yong-Bin¹,Zhang Zhuo^1,2

Online:2015-05-04 Published:2015-05-04
About author:

摘要/Abstract

摘要： 针对K一均值聚类算法存在的缺陷，将改进的粒子群优化算法—智能单粒子优化算法(lSPO)应用到聚类分析当中来，提出一种混合聚类算法ISPO}K-means.该算法分为两个阶段:第一阶
段利用ISPO算法较强的全局寻优能力形成初始聚类，第二阶段将初始聚类结果通过K-mean、算法形成最终聚类结果输出.与K一均值聚类算法和基于传统粒子群K一均值混合算法进行比较，在低维、高
维和多样木高维三种特征数据集上实验，结果表明提出的算法能够有效克服其它算法易陷入局部最优的问题，并且验证了算法在对多样木高维数据集处理上具有较明显优势.

Abstract: After analyzing the disadvantages of the classical K-means clustering algorithm, an improved particle swarm optimization algorithm-intelligent single particle optimizer(ISPO)，is applied to cluster analysis, and this
paper proposes a novel hybrid K-means clustering based on lSPO. The hybrid ISPO+K-means algorithm includes two modules, the lSPO module and K-means module. At the initial stage, the lSPO module is executed for a short
period to search for the initial clusters’results.The clusters’results arc transferred to the K-means module for refining and generating the final optimal clustering solution. Comparison of the performance of the proposed
approach with the cluster method based on K-means and traditional PSO+K-means algorithm is experimented. The experimental results show the proposed method can effectively solve the premature convergence problem, and
verified the algorithm for variety of high-dimensional data sets processing has obviously the most advantageous.

陈永彬1，张琢1,2**
. 智能单粒子优化算法在聚类分析中的应用*
[J]. 南京大学学报(自然科学版), 2011, 47(5): 578–584.

Chen Yong-Bin¹,Zhang Zhuo^1,2. An application of intelligent single particle optimizer in cluster analysis
[J]. Journal of Nanjing University(Natural Sciences), 2011, 47(5): 578–584.

参考文献

[1]Ghao H J，Zhong C M, Li W, et al. Clustering and visualization of web search results. Journal of Nanjing University(Natural Sciences)，2010,6(5):542-544.(赵华军，钟才明，李文等.网页搜索结果聚类与可视化.南京大学学报(自然科学),2010,46(5):542-544).
[2]Berkhin P. Survey of clustering data mining techniques. San Josc:Accrue Software，2002
[3]Tou J T,Gonzalez R C.Pattern recognition principle. Addison Wesley, 1974，377.
[4]Omran M, Salman A，Engclbrecht A P. image Classification using Particle Swarm Optimiza- tion. Conference on Simulated Evolution and Learning, 2002，1:370一374.
[5]Gao S, Yang J Y. New clustering method based on particle swarm algorithm. Journal of Nanjing University of Aeronautics and Astronautics，2006, 38(B07); 62-65.(高尚，杨静宇.一种新的基于粒子群算法的聚类方法.南京航空航天大学学报，2006, 38(B07): 62-65).
[6]Liu J M，H an I. C. Cluster analysis based on particle swarm optimization algorithm. Systems Engineering-Theory and Practice，200，6:54 -58.(刘靖明，韩丽川.基于粒子群的K一均值聚类算法.系统工程理论与实践，2005, 6: 54一58).
[7]Ding X Q, Jiao S M. PSO spatial clustering with obstacles constraints，Computer Engineer- ing and Design, 2007，28(24):5924一5926. (丁晓晴，焦素敏.基于粒子群优化的带障碍约束空间聚类分析.计算机工程与设计，2007, 28(24):5924一5926).
[8]Zhang C S, Sun J G, Yang F Q. Research on a pso based dynamic clustering algorithm. Journal of Computer Research and Development，2007，44:89-93.(张长胜，孙吉贵，杨凤芹.一种基于PSO的动态聚类算法.计算机研究与发展，2007，44:89一93).
[9]Pei Z K，Hua X. The clustering algorithm based on particle swarm optimization algorithm. international Conference on intelligent Compu- tation Technology and Automation,2008，148一 l5l.
[10]Cui X, Potok T E. Document clustcring analy sis based on hybrid PSO+K-means algorithm Journal of Computer Sciences，Special lssue 2005，4:27一33.
[11]Zhan Z H，Zhang J，Li Y，et al. Adaptive par- tice swarm optimization, IEEE Transactions on Systems, Man, and Cybernetics Part B, 2009，39:1362一1381.
[12]Chen J Y. Research on amelioration of cluste- ring algorithm based on pso for the application of web usage pattern. Master thesis. Tianjin; Tianjin University, 2007, 21 ~24.陈君彦.基
于粒子群的聚类算法改进及其在访问模式中的应用研究.硕士论文.天津大学，2007,21一 24)
[13]Gao S, Yang J Y. Swarm intelligence algo- rithms and applications. Beijing; China Water Power Press, 2006 , 100~104.(高尚，杨静宇.群智能算法及其应用.北京:中国水利水电出版社，2006，100一104).
[14]Ji Z, Zhou J R，Liao H L,etal. A novel intclli- gent single particle optimizer. Chinese Journal of Computers, 2010, 33(3):556一560.(纪震，周家锐，廖惠莲等.智能单粒子优化算法. 计算机学报，2010, 33(3): 556-560).
[15]Wang C J，Dong X G, Diao X W. Cluster anal- ysis based on particle swarm optimization with immunity. Journal of Uuangxi Normal Univcrsi- ty; Natural Science Edition, 2008，26(3):166一
167.(土纯杰，董小刚，刁心薇.基于免疫粒子群的K均值聚类算法.]’一西师范大学学报(自然科学)，2008, 26(3): 166-167).
[16]Sun Y，Luo K. Clustering method based on proved particle swarm optimization. Computer Engineering and Application, 2009，45(33): 133-134.(孙洋，罗可.基于改进粒子群算法的聚类算法.计算机工程与应用，2009, 45(33):133一134).
[17]Long H X，Xu W B, Sun J. Data clustering based on quantum-behaved particle swarm opti mization. Application Research of Computers 2006, 12; 4l~42.(龙海侠，须文波，孙俊. 基于QPSO的数据聚类.计算机应用研究 2006，12:4l一4 2 ).
[18]UCI Machine Learning Repository; Data Sets.http://archive.ics.uci.edu.ml/datasets.html.2010-09

im〕

Metrics

Viewed

Full text

Abstract

Cited

Shared

Discussed