基于自组织特征映射(SOM)网络对潜在客户的挖掘

通过自组织特征映射网络（SOM）帮助市场分析人员对消费者的消费记录进行分析,从而概括出每一类消费者的消费模式。实现对消费群体的区分和对潜在客户的挖掘。SOM神经网络是一种基于模型的聚类方法。该网络通过对样本数据实施标准化、规范化的规定,将每条样本数据看作为一个基因数据,通过对基因数据的聚类,找出功能相似的基因。从而达到对潜在客户的挖掘。抽取一定量的样本数据,通过Matlab平台建立一个SOM人工神经网络对数据进行训练,结果可以发现当训练步骤达到一定量的时候,样本数据向不同的“中心点”聚集,最后达到聚类的效果,从而发现一些容易被忽视的点,也就是潜在客户。

Application of Discovering the Potential Customers Based on Self-Organizing Feature Map(SOM)Neural Network

It tends to help the analyzers to analyze the consumption record of the customers by this network so as to summarize the consumption mode of every kind of customers and to realize the discovery of potential customers.Self-Organizing Feature Map(SOM) neural network is a kind of clustering methodology based on modeling.The network treat each sample as a piece of genetic data through standardizing each of them,clustering is a fundamental discovery application of human being,by clustering the genetic data to find out the familiar gene.It took amount of samples,built an SOM neural network through Matlab platform for training the data and it could be found when the training steps came to some level the samples will cluster to a different "center spot" which could be ignored normally and that was the potential customers.