[1]吴永芬,冯茂岩,张 健.基于小生境粒子群的属性约简算法[J].南京师范大学学报(工程技术版),2008,08(04):132-135.
 Wu Yongfen,Feng Maoyan,Zhang Jian.Attribute Reduction Algorithm Based on the Niche PSO[J].Journal of Nanjing Normal University(Engineering and Technology),2008,08(04):132-135.
点击复制

基于小生境粒子群的属性约简算法
分享到:

南京师范大学学报(工程技术版)[ISSN:1006-6977/CN:61-1281/TN]

卷:
08卷
期数:
2008年04期
页码:
132-135
栏目:
出版日期:
2008-12-30

文章信息/Info

Title:
Attribute Reduction Algorithm Based on the Niche PSO
作者:
吴永芬1 冯茂岩2 张  健3
1. 解放军理工大学指挥自动化学院, 江苏南京210007; 2. 江苏海事职业技术学院信息工程系, 江苏南京210070;
3. 三江学院计算机基础部, 江苏南京210012
Author(s):
Wu Yongfen1Feng Maoyan2Zhang Jian3
1.Institute of Command Automation,PLA University of Science and Technology,Nanjing 210007,China;2.Department of Information Engineering,Jiangsu Maritime Institute,Nanjing 210070,China;3.Department of Computer Elementary Training,Sanjiang University,Nanjin
关键词:
小生境 粒子群算法 粗糙集 属性约简
Keywords:
n iche PSO rough set attribute reduction
分类号:
TP301.6
摘要:
遗传算法(GA)及蚂蚁算法(ACO)等进化属性约简算法,具有全局寻优的优点,但存在算法时间复杂度高,搜索空间大等不足;粒子群(PSO)属性约简算法,虽然可提高求解效率,但易陷入局部最优.本文引入小生境技术,提出基于小生境粒子群的属性约简算法,利用小生境技术造就种群的多样性,使解保持多样化,以此避免粒子群属性约简算法易早熟收敛的缺点.理论分析及实验结果表明,该算法是有效可行的.
Abstract:
Curren tly, there arem any evolutionary a lgo rithm s ava ilab le fo r a ttr ibute reduction as genetic a lgor ithm ( GA ) and the ant co lony optim iza tion a lgor ithm ( ACO), but h igh tim e com plex ity and w ide search space are the ir comm on d isadvantages. The PSO attribute reduction a lgo rithm ough t to improve the e ffic iency, how eve r, it has the prem ature conv ergence problem. Therefore, an attribute reduc tion a lgor ithm based on the n iche PSO is presented in this paper a fter introduc ing the niche techno logy. The n iche techno logy is app lied to m ainta in the diversity of the popu lation and the solu tions, and avo id the prem ature converg ence problem o f PSO a lgor ithm. The theoretica l analysis and exper imenta l resu lts ind icate tha t the proposed niche PSO a lgo rithm is feasible and effec tive.

参考文献/References:

[ 1] W ang Jue, M iao Duoq ian. Ana ly sis on attr ibute reduction strateg ies o f rough set [ J]. Journal o f Com pute r Science and Technology, 1998, 13( 2): 189-193.
[ 2] 徐章艳, 刘作鹏, 杨炳儒, 等. 一个复杂度为m ax( O ( |C | |U | ), O ( |C |2 |U /C | ) )的快速属性约简算法[ J] . 计算机学报, 2006, 29( 3) : 391-399.
Xu Zhangyan, L iu Zuopeng, Yang B ing ru, e t a.l A qu ick attribute reduction a lgo rithm w ith com plex ity o fm ax( O( |C | |U | ),O ( |C |2 |U /C | ) ) [ J]. Computer Journa,l 2006, 29( 3): 391-399. ( in Chinese)
[ 3] Oh I S, Lee J S, M oon B R, et a.l H ybr id g enetic a lgo rithm s fo r feature se lection[ J] . IEEE Transac tions on Pattern Analysis and Machine Inte lligence, 2004, 26( 1): 1 424-1 437.
[ 4] Jiang Yuanchun, L iu Ye zheng. An a ttr ibu te reduction m ethod based on ant co lony optim iza tion[ C ] / / Proceedings of the 6 th World Cong ress on In tellig ent Contro l and Autom ation. W ashing ton: IEEE Com pute r Society Press, 2005: 3 542-3 546.
[ 5] Dai Jianhua, ChenW e idong, GuoH ongy ing, e t a .l Partic le sw arm algor ithm fo rm in im al attribute reduction of dec is ion data tab
les[ C ] / / Pro cessing of the 1st Inte rnationalM ult-i symposium s on Com puter and Com putationa l Sc ience. W ash ing ton: IEEE Com puter So ciety Press, 2006: 3 021-3 025.
[ 6] 王杨. 基于小生境遗传算法的粗糙集属性约简方法[ J] . 计算机工程, 2008, 34( 5): 66-70.
W ang Yang. Rough set attribute reduc tion a lgor ithm based on n icheGA[ J]. Com puter Eng ineer ing, 2008, 34( 5) : 66-70. ( in Chinese)
[ 7] W ang Guoy in, Zhao Jun, Jiu Jiang, et a.l Theo re tica l study on attribute reduction o f rough se t theory: comparison of algebra
and inform ation v iew s[ C ] / / Proceed ings o f the 3rd IEEE Interna tiona l Conference on Cogn itive Inform atics. W ash ington:IEEE Com pute r Society Press, 2004: 148-155.
[ 8] Lee C G, Cho D H, JungH K. N iche gene tic a lgor ithm w ith restricted competition se lection form ultim oda l function optim ization[J]. IEEE T rans onM agnetics, 1999, 35( 3): 1 122-1 125.

备注/Memo

备注/Memo:
通讯联系人: 吴永芬, 硕士, 研究方向: 粗集理论与应用. E-m ail: yfw u0916@ 126. com
更新日期/Last Update: 2013-04-24