| 基于进化树型自组织神经网络聚类分析 |
| |
| 引用本文： | 沈来信,杨帆.基于进化树型自组织神经网络聚类分析[J].黄山学院学报,2007,9(3):38-40. |
| |
| 作者姓名： | 沈来信 杨帆 |
| |
| 作者单位： | 黄山学院,信息工程学院,安徽,黄山,245021 |
| |
| 摘 要： | SOM是当前最著名的基于神经网络的用于数据可视化、聚类等任务的数据分析工具之一,为了克服传统SOM模型需要预选指定的限制,特别是在大的映射网络中寻找最佳匹配结点将会很耗时,所以提出一种新的进化树型自组织神经网络ET-SOM。通过对人工数据和工业数据进行实验证实了模型的有效性。
|
| 关 键 词： | 数据可视化 聚类 最佳匹配结点 进化树型自组织神经网络 |
| 文章编号： | 1672-447X（2007）03-0038-03 |
| 修稿时间： | 2007-03-22 |
Clustering Analysis Based on Evolutionary Tree-Structured SOM Network |
| |
| Shen Laixin,Yang Fan.Clustering Analysis Based on Evolutionary Tree-Structured SOM Network[J].Journal of Huangshan University,2007,9(3):38-40. |
| |
| Authors: | Shen Laixin Yang Fan |
| |
| Institution: | Information and Engineering College, Huangshan University, Huangshan245021, China |
| |
| Abstract: | The Self-Organizing Map is a very well-known data analysis tool for tasks like data visualization and clustering.To overcome the restriction that the tradition SOM model must designate in advance,especially the time-consuming search for the Best Matching Unit in large maps,we propose a new Evolutionary Tree-Structurod Self-Organizing Maps.Test results with both synthetic and actual data show that the model works quite well. |
| |
| Keywords: | data visualization clustering BMU Evolutionary Tree-Structured Self-Organizing Maps |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
