|
|||||
|
|
| 一种新的基于混合蚁群算法的聚类方法 | |
| 引用本文： | 高尚,汤可宗,杨静宇.一种新的基于混合蚁群算法的聚类方法[J].微电子学与计算机,2006,23(12):38-40,43. |
| 作者姓名： | 高尚 汤可宗 杨静宇 |
| 作者单位： | 1. 江苏科技大学,电子信息学院,江苏,镇江,212003;苏州大学江苏省计算机信息处理技术重点实验室,江苏,苏州,215006 2. 江苏科技大学,电子信息学院,江苏,镇江,212003 3. 南京理工大学,计算机科学与技术系,江苏,南京,210094 |
| 摘 要： | 建立了聚类分析问题模型，分析了K-均值算法、模拟退火算法和基本蚁群算法的优缺点。对蚁群算法作了改进．思路是K-均值方法混合，利用K-均值方法的结果作为初值。经过比较测试，两种混合蚁群算法的效果都比较好．特别混合方法二的效果最好。 |
| 关 键 词： | 聚类分析 蚁群算法 K-均值算法 模拟退火算法 |
| 文章编号： | 1000-7180（2006）12-0038-03 |
| 收稿时间： | 2005-11-21 |
| 修稿时间： | 2005-11-21 |
A New Clustering Algorithm Based on Hybrid Ant Colony Algorithm |
|
| GAO Shang,TANG Ke-zong,YANG Jing-yu.A New Clustering Algorithm Based on Hybrid Ant Colony Algorithm[J].Microelectronics & Computer,2006,23(12):38-40,43. | |
| Authors: | GAO Shang TANG Ke-zong YANG Jing-yu |
| Affiliation: | 1 School of Electronics and Information, Jiangsu University of Science and Technology, Zhenjiang 212003,China;2 Provincial Key Lab. of Computer Infor. Processing Technology, Soochow University, Suzhou 215006, China;3 Department of Computer Sci. and Tech., Nanjing University of Science and Technology, Nanjing 210094, China |
| Abstract: | An optimization model of clustering problem is given in this paper. The advantages and shortages of K- Means algorithm, simulated annealing algorithm and basic ant colony algorithm are analyzed. The algorithm is then extended to use K-means clustering to seed the initial solution and the information pheromone is adjusted according to them. All the 2 hybrid ant colony algorithms are proved effective and especially the second hybrid algorithm is a best algorithm than others. |
| Keywords: | Clustering analysis Ant colony algorithm K-Means algorithm Simulated annealing algorithm |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ | |