基于关联规则的决策树算法

Decision Tree Algorithm Based on Association Rules

通过将关联规则与决策树算法相结合,形成一种基于关联规则的决策树算法。该算法对不同时期同一事务的异种数据结构进行处理,得到一种可扩展的多分支分类决策树,使得改进后的决策树算法具有良好的可扩展性。该算法解决了传统分类算法在数据集维度发生变化时分类过程无法持续进行的问题。

This paper combines association rules and decision tree algorithm,and proposes a new decision tree classification based on association rule.The decision tree algorithm can handle dissimilar transaction data set record blocks which are same investigations conducted in different times to the same transactions.Through the decision tree algorithm,it can get a multi-crunodes decision tree,which has a good extendable performance.The algorithm solves the problem,which exists in the traditional classification,that is the traditional classification can not classify effectively and sustaine when dimensions of dataset change.