| 基于蚁群算法的多字符聚类识别 |
| |
| 引用本文： | 王立群,杨淑莹,安博.基于蚁群算法的多字符聚类识别[J].天津理工大学学报,2008,24(5). |
| |
| 作者姓名： | 王立群 杨淑莹 安博 |
| |
| 作者单位： | 天津理工大学,计算机科学与技术学院,天津,300191 |
| |
| 基金项目： | 天津市自然科学基金，天津市教委资助项目 |
| |
| 摘 要： | 本文采用蚁群算法对聚类数目已知的多字符进行聚类识别,在分析了基本蚁群算法的基础上,提出了一种改进的蚁群算法,该算法结合分布式计算、正反馈机制、贪婪式搜索算法等.对每只蚂蚁构造一个可行解,利用信息素矩阵,经过若干次的选代,找寻包含最优解的蚂蚁.通过与K-means和遗传算法比较,最后得出结论,该蚁群算法识别效果好,执行效率高.
|
| 关 键 词： | 聚类识别 蚁群算法 多字符 |
Multi-character cluster recognition based on ant algorithm |
| |
| WANG Li-qun,YANG Shu-ying,AN Bo.Multi-character cluster recognition based on ant algorithm[J].Journal of Tianjin University of Technology,2008,24(5). |
| |
| Authors: | WANG Li-qun YANG Shu-ying AN Bo |
| |
| Abstract: | An improved ant algorithm based on class number-known multi-character cluster recognition problem has been pro- posed in this paper,which has combined distributed computing,positive feedback mechanisms and greed search algorithm. For each ant,a possible solution is constructed and pheromone matrix is used to look for ant including optimal solution.Com- pared with K-means and GA,a conclusion is drawn that the improved ant algorithm is efficient and ideal. |
| |
| Keywords: | ant algorithm multi-character cluster recognition |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
