论文部分内容阅读
概念层次在数据挖掘中有着重要的作用 .通过自动生成概念层次 ,可有效地提高数据挖掘的效率 ,在不同层次上发现知识 .文中介绍基于云模型的数值型概念表示方法 ,通过云模型的期望值、熵和超熵三个数字特征有效地表达定性概念 ,并实现定性和定量的不确定转换 .通过云变换实现了泛概念树中叶结点的自动生成 ,并自动构造数值型数据的泛概念树 .同时 ,进一步研究了泛概念树中的概念爬升和跳跃的方法 ,为通过数据挖掘发现各层次知识提供了基础 .
Conceptual hierarchy plays an important role in data mining.Automatic generation of conceptual hierarchy can effectively improve the efficiency of data mining and find knowledge at different levels.In this paper, we introduce a numerical concept representation method based on cloud model, through the expectation of cloud model , Entropy and Superextrue, which can express the qualitative concept effectively and realize the qualitative and quantitative uncertainty transformation.Firstly, by means of cloud transformation, the automatic generation of leaf nodes in the universal conceptual tree is realized, and a generic concept tree of numerical data is constructed automatically At the same time, we further study the concepts of concept climbing and jumping in pan-concept tree, which provides the basis for discovering knowledge at different levels through data mining.