A Classification Method Based on Subspace Clustering and Association Rules
A Classification Method Based on Subspace Clustering and Association Rules
复制标题
DOI：
10.1007/s00354-007-0015-7
复制
发表时间：
2007
影响因子：
2.6
通讯作者：
T. Washio;Koutarou Nakanishi;H. Motoda
中科院分区：
文献类型：
--
作者：
T. Washio;Koutarou Nakanishi;H. Motoda
基于类别关联规则（CAR）的分类因其高度的可解释性和准确性而成为近年来数据挖掘研究中日益热门的话题。然而，大多数的方法没有集中解决的分类实例，包括数值属性。本文提出了一种基于超矩形聚类的层次子空间聚类方法，以有效地提供定量、解释性和准确的汽车。通过对UCI存储库数据的测试，证明了所提出的方法的显着性能。
Class Association Rule (CAR) based classification is a growing topic in recent datamining study for its high interpretability and accuracy. However, most of the approaches have not intensively addressed the classification of instances including numeric attributes. In this paper, a levelwise subspace clustering deriving hyper-rectangular clusters is proposed to efficiently provide quantitative, interpretative and accurate CARs. Significant performance of the proposed approach has been demonstrated through the tests on UCI repository data.