基于自组织映射和生物地理优化的聚类算法A Mixed SOM and Biogeography-based Optimization Algorithm for Clustering
温肖谦;黄发良;李超雄;汪焱;
摘要(Abstract):
针对传统生物地理优化算法(bio-geographic optimization algorithm,BBO)的种群随机初始策略会降低聚类算法性能的问题,提出了一种基于自组织映射算法(self-organization feature map,SOM)和BBO的混合聚类算法(improved SOM and bio-geography optimization,ISOMBBO),通过优化初始化神经元权值的方法改进SOM算法,然后以改进的SOM来计算数据聚类的初始簇中心,最后在BBO优化框架下进行数据簇结构的寻优.在4个标准数据集(Iris、Wine、Glass与Diabetes)的实验中,实验结果表明该算法不仅提高聚类的有效性,而且相对于传统的优化算法具有更好的优化能力和收敛度.
关键词(KeyWords): BBO算法;聚类;SOM算法;优化
基金项目(Foundation): 教育部人文社会科学研究青年基金资助项目(12YJCZH074);; 福建省教育厅资助项目(JA13077)
作者(Authors): 温肖谦;黄发良;李超雄;汪焱;
参考文献(References):
- [1]周涛,陆惠玲.数据挖掘中聚类算法研究进展[J].计算机工程与应用,2012,48(12):100-111.
- [2]Simon D.Biogeography-based optimization[J].IEEE Transaction on Evolutionary Computation,2008,12(6):702-713.
- [3]Vijay Kumar,Jitender Kumar Chhabra,Dinesh Kumar.Advances in computing communication and control[M].Berlin:Springer Berlin Heidelberg,2011:448-456.
- [4]Gong W,Cai Z,Ling C X.DE/BBO:a hybrid differential evolution with biogeography-based optimization for global numerical optimization[J].Soft Computing,2010,15(4):645-665.
- [5]Ma Haiping.An analysis of the equilibrium of migration models for biogeography-based optimization[J].Information Sciences,2010,180(18):3444-3464.
- [6]林剑,徐力.基于混合生物地理优化的混沌系统参数估计[J].物理学报,2013,62(3):1-7.
- [7]Boussad I,Chatterjee A,Siarry P.Two-stage update biogeography-based optimization using differential evolution algorithm(DBBO)[J].Computers and Operations Research,2011,38(8):1188-1198.
- [8]吕强,俞金寿.基于粒子群优化的自组织特征映射神经网络及应用[J].控制与决策,2005,20(10):1115-1119.
- [9]丁硕,常晓恒,巫庆辉.基于自组织特征映射神经网络的聚类分析[J].信息技术,2014(6):18-21.
- [10]谭维,杨燕.基于自组织特征映射的聚类集成算法[J].计算机工程与设计,2010,31(22):4885-4888.
- [11]Sugiyama A,Kotani M.Analysis of gene expression data by using self-organizing maps and K-means clustering[J].Neural Network,2002,5(5):1342-1345.
- [12]Xiao X E.Gene clustering using self-organizing maps and particle swarm optimization[J].International Parallel and Distributed Processing Symposium,2003,11(4):154-163.