摘要
研究了一种基于自组织映射(Self-Organizing Map,SOM)神经网络的交易数据库聚类方法,该方法首先对数据库中的数据项进行SOM训练学习产生初步的聚类结果,然后对第一次获得的聚类结果进行二次聚类,与直接聚类方法相比,该方法提高了聚类的效率,减少了计算时间。
This paper studies a clustering algorithm of transaction database based on serf-organizing map（SOM） neural network. The method first uses SOM to produce a clustering restflt by learning data item of transaction database. Then, in the second stage, the clustering results are clustered again. To compare with other clustering method for transaction database, the method performs well and reduces runtime.
出处
《计算机与现代化》
2009年第12期36-38,171,共4页
Computer and Modernization
