|本期目录/Table of Contents|

[1]刘世华,叶展翔,刘向华,等.一种新的自组织粒子群聚类算法[J].温州职业技术学院学报,2015,03:54-58.
 LIU Shihua,YE Zhanxiang,LIU Xianghua.A New Self-Organized PSO Clustering Algorithm[J].Journal of Wenzhou Vocational and Technical College,2015,03:54-58.
点击复制

一种新的自组织粒子群聚类算法(PDF)

《温州职业技术学院学报》[ISSN:1006-6977/CN:61-1281/TN]

期数:
2015年03期
页码:
54-58
栏目:
应用技术
出版日期:
2015-08-20

文章信息/Info

Title:
A New Self-Organized PSO Clustering Algorithm
作者:
刘世华;叶展翔;刘向华;
温州职业技术学院信息技术系;
Author(s):
LIU Shihua YE Zhanxiang LIU Xianghua
Information Technology Department, Wenzhou Vocational & Technical College, Wenzhou, 325035, China
关键词:
粒子群聚类竞争学习PSOSOMSOPSC
Keywords:
PSO clustering Competition learning PSO SOM SOPSC
分类号:
TP311.13
DOI:
10.13669/j.cnki.33-1276/z.2015.059
文献标识码:
A
摘要:
针对粒子群优化(PSO)算法复杂度偏高的问题,提出一种新的基于竞争学习的自组织粒子群聚类(SOPSC)算法。该算法采用每个粒子代表一个聚类中心的编码方法,通过借鉴自组织映射(SOM)算法的竞争学习机制,采用类内相似度和类间相异度作为指导,使粒子进行自组织飞行,从而达到自动聚类的目的,克服了传统粒子群聚类算法中粒子编码复杂、算法复杂度偏高的缺点。实验证明,该算法聚类精度高、稳定性好,且对初始值和参数不敏感。更多还原
Abstract:
A Self-organized PSO Clustering (SOPSC) algorithm based on competition learning was proposed against the complexity of PSO clustering algorithm. The algorithm coded one particle as a center of each cluster for clustering analysis. Through the competition learning of SOM algorithm, particles flew as self-organized manner directed by the inner similarity and dissimilarity between different clusters to achieve the goal of self-clustering. It overcame the complexity of particle coding and algorithm of traditional PSO clustering. It was tested that the algo?rithm improves the cluster accuracy and its stability and is not sensitive to the initial value and parameters.

参考文献/References

-

备注/Memo

备注/Memo:
-
更新日期/Last Update: 2015-09-20