| 动态聚类的一种新方法——最小组内平方和法 |
| |
| 引用本文： | 顾世梁,莫惠栋.动态聚类的一种新方法——最小组内平方和法[J].扬州大学学报(农业与生命科学版),1989(4). |
| |
| 作者姓名： | 顾世梁 莫惠栋 |
| |
| 作者单位： | Gu Shiliang;Mo Huidong Laboratary of Quantitative Genetics，Jiangsu Agricultural College，Yangzhou |
| |
| 摘 要： | 本文提出一种以组内平方和之和最小为标准的新的动态聚类方法,简称为最小组内平方和法(MinSS_w法).Monte-Carlo试验表明,该法能有效地调整初始分组中的个体.使之达到合理的分类,并有着良好的稳健性;而目前常用的重心法(K-means法)的动态聚类结果,却是相当不稳定的.上述结论也为一个含14个小麦品种的聚类实例所证实.
|
| 关 键 词： | 聚类分析 动态聚类 最小组内平方和 |
A New Dynamic Clustering Method-MinSSw Method |
| |
| Gu Shiliang,Mo Huidong Laboratary of Quantitative Genetics,Jiangsu Agricultural College,Yangzhou.A New Dynamic Clustering Method-MinSSw Method[J].Journal of Yangzhou University:Agricultural and Life Science Edition,1989(4). |
| |
| Authors: | Gu Shiliang Mo Huidong Laboratary of Quantitative Genetics Jiangsu Agricultural College Yangzhou |
| |
| Abstract: | A new dynamic clustering method which minimizes the total sum of squares within groups, the MinSS_W method, along with its algorithm was developed. The Monte-Carlo sampling was used to investigate the properties of the new method and to make some comparisons with K-means method. The new method can adjust entries in ini- tial groups into proper placement effectively. The performance of the new method is much more robust than that or K-means method in all situations. This is also demonstrated by a real example of partitioning 14 wheat varietics into 5 groups. |
| |
| Keywords: | cluster ana1ysis dynamic clustering minimum sum of squares within groups |
| 本文献已被 CNKI 等数据库收录！ |
