计算机工程与应用 ›› 2021, Vol. 57 ›› Issue (3): 87-93.DOI: 10.3778/j.issn.1002-8331.1911-0242

• 理论与研发 • 上一篇    下一篇

粗糙集属性依赖度强化的应急数据挖掘模型

高天宇,王庆荣,杨磊   

  1. 兰州交通大学 电子与信息工程学院,兰州 730070
  • 出版日期:2021-02-01 发布日期:2021-01-29

Data Mining Model Based on Attribute Dependability Enhancement of Rough Set

GAO Tianyu, WANG Qingrong, YANG Lei   

  1. School of Electronic and Information Engineering, Lanzhou Jiaotong University, Lanzhou 730070, China
  • Online:2021-02-01 Published:2021-01-29

摘要:

在粗糙集的核心方法属性约简理论中,针对次要属性过多时属性依赖度一致引发的约简困难现象,以粗糙集属性依赖度强化为基本方法构建数据挖掘模型。模型中,计算各属性在组合中的依赖度,将组合中的依赖度合并于各属性从而强化属性依赖度。为合理化计算过程,给出一种离散化方法并构建探索模型进行实验,缩小离散量范围与属性组合范围。最后所得依赖度作为属性重要性,挖掘重要属性。实验证明,在部分地区5级以上震后经济损失影响因素分析中,比较传统属性约简,该模型可更有效地分析属性之间的重要性,解决了约简困难的问题,使用测试数据时该模型与传统方法的结果有86%的一致性。

关键词: 数据挖掘, 粗糙集, 属性约简, 强化属性依赖度, 粒度, 经济损失

Abstract:

As the core of rough set, the attribute reduction theory is difficult to make a reduction when there are too many secondary attributes. Based on the enhanced attribute dependability of rough set, a data mining model is constructed. In this model, the dependability of each attribute in group is calculated, and the dependability in combination is merged into each attribute to enhance the attribute dependability. For the rational of exploration model calculation process, a exploration model and a discretization method are built. The range of discrete quantity and attribute combination is narrowed. As the value of attributes, the obtained dependability is used to find the important attributes. The experimental results show that compared with the traditional attribute reduction, this model can more effectively analyze the importance of attributes and solve the difficult problem of reduction in the analysis of factors affecting the economic loss after the earthquake of magnitude 5 or above in some areas. The result of this model in test data is 86% consistent with that of the traditional method.

Key words: data mining, rough set, attribute reduction, dependability enhancement, granularity, economic loss