| 基于佳点集遗传算法的聚类技术 |
| |
| 引用本文： | 苏守宝,刘仁金. 基于佳点集遗传算法的聚类技术[J]. 计算机应用, 2005, 25(3): 643-645 |
| |
| 作者姓名： | 苏守宝 刘仁金 |
| |
| 作者单位： | 皖西学院,计算机科学与技术系,安徽,六安,237012;安徽大学,计算机科学与技术学院,安徽,合肥,230039;皖西学院,计算机科学与技术系,安徽,六安,237012;安徽大学,计算机科学与技术学院,安徽,合肥,230039 |
| |
| 摘 要： | 提出了两种基于佳点集遗传算法的聚类新方法GAmeans和HgaMeans，适用于不同数据库下的聚类挖掘。GAmeans可用于发现指定簇数的聚类中心，具有对初始数据的弱依赖性、收敛快、精度高并可避免早熟的特点；而混合方法HgaMeans是利用k-means对GAmeans聚类结果的进一步提炼，实验表明它具有更好的聚类质量和综合性能。
|
| 关 键 词： | 聚类 佳点集 遗传算法 |
| 文章编号： | 1001-9081(2005)03-0643-03 |
Clustering techniques based on good point-set genetic algorithm |
| |
| SU Shou-bao,LIU Ren-jin. Clustering techniques based on good point-set genetic algorithm[J]. Journal of Computer Applications, 2005, 25(3): 643-645 |
| |
| Authors: | SU Shou-bao LIU Ren-jin |
| |
| Affiliation: | SU Shou-bao~ |
| |
| Abstract: | This paper proposed two new approaches, GAmeans and HgaMeans, using good point-set genetic algorithm to cluster data in data mining for various databases. GAmeans could be used to find the centroid of specified number of clusters, characterized as inferior sensitivity to initial data, quick convergence, high accuracy and removal of premature. The hybrid algorithm HgaMeans used k-means to refine the clusters formed by GAmeans. Experiments show that the hybrid method has better clustering quality and general performance. |
| |
| Keywords: | clustering good point-set genetic algorithm |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
