连续数据的分割及关联规则发现

Association Rule Mining from Interval of Continuously Valued Data

关联规则的挖掘是一个重要的数据挖掘问题，目前的算法主要是研究基于支持－信任框架理论的关联规则挖掘，但是基于支持－信任框架理论的关联规则只适用于交易类型的数据库，然而现实的数据库中有许多连续数据，经典的关联规则就不适用了．该文介绍一种对连续数据集进行须处理过程，即对数据库中的数据项进行距离划分，并给出基于聚类方法的算法设计思想．

Mining association rules is an important data mining problem. There have been many algorithms Droposed to efficicntly discover association rules from database. However algorithms based on support-confidence framework have some weakness. This paper introduces a kind of approach to process continuous valued data. that is interval data based on distance. An algorithm based on clustering approach is given so that the effectiveness of the mined knowledge can be improved.