基于模糊均值聚类和感知机的网络银行客户挖掘

通过结合运用模糊均值聚类和感知机进行网络银行目标营销,首先利用模糊c均值方法(FCM)找出频繁使用网络银行服务的客户,再进一步用这些客户样本对一个多层感知机(MLP)进行训练,最后利用训练好的感知机对银行的非网络银行注册客户进行分类,找出潜在的网上银行服务频繁使用者,银行即可对这些客户进行重点推广网络银行服务。

Internet Bank Customer Mining Based on Fuzzy c-means Clustering and Perceptron

This paper presents a data mining approach combining the fuzzy c-means and perceptron for target marketing in Internet bank. To begin with, it is necessary to utilize the fuzzy c-means method to find frequent Internet bank users as samples. Then we train the multiple layers perceptron by these samples. In the end, we use the trained perceptron to classify the bank non-registered customers , When potential frequent Internet bank users are thus identified, target marketing can be conducted to these customers.