Abstract:

Self generating neural networks have been presented as a better alternative to fixed structure in data mining application. The paper provided the algorithm of Self Organising Map(SOM), and presented an version of Growing Self Organising Map(GSOM) model which can extend into a nearest prototype classifier, and gave an application in data mining using GSOM.

Key words: Self organizing map, Clustering, Nearest prototype

摘要：

在数据挖掘应用中,基于自生成神经网络的方法被认为是比基于固定网络方法更好的一种替代方法。介绍了自组织映射(SOM)算法和生长的自组织映射 (GSOM)模型,证明了GSOM的功能可以扩展成最近原型分类,并给出了其在数据挖掘中的一个应用。

关键词: 自组织映射, 聚类, 最近原型