基于聚类和决策树的用户分层研究

Study on User Stratification Based on Clustering and Decision Tree

为解决大数据背景下众多行业面临的用户分层问题,以主成分分析(PCA)方法对聚类模型的用户分层结果进行优劣比较,然后使用决策树进行事后分类,通过提取决策树的节点,给出有明确业务含义的分类标准,利用这些简单标准做更快速的分层估计。

In order to solve the problem of user stratification faced by many industries in the context of big data,this paper uses PCA to compare the advantages and disadvantages of the user stratification results of the clustering model,and then uses the decision tree for post-event classification.By extracting the nodes of the decision tree,it is given There are classification standards with clear business meanings,and these simple standards can be used to make faster hierarchical estimation.