计算机工程与应用 ›› 2018, Vol. 54 ›› Issue (14): 89-94.DOI: 10.3778/j.issn.1002-8331.1703-0422

• 大数据与云计算 • 上一篇    下一篇

一种改进的邻域多粒度粗糙集模型

刘  丹1,徐立新1,孟慧丽2,朱命冬1   

  1. 1.河南工学院 计算机科学与技术系,河南 新乡 453003
    2.河南师范大学 计算机与信息工程学院,河南 新乡 453007
  • 出版日期:2018-07-15 发布日期:2018-08-06

Improved neighborhood multi-granulation rough set model

LIU Dan1, XU Lixin1, MENG Huili2, ZHU Mingdong1   

  1. 1.Department of Computer Science & Technology, Henan Institute of Technology, Xinxiang, Henan 453003, China
    2.College of Computer and Information Engineering, Henan Normal University, Xinxiang, Henan 453007, China
  • Online:2018-07-15 Published:2018-08-06

摘要: 目前,邻域多粒度粗糙集模型广泛采用的距离函数闵可夫斯基距离存在着一定的局限性,通过引入兰氏距离作为距离函数,重构了邻域半径的选取方法,基于此提出一种改进的邻域多粒度粗糙集模型,并证明了相关的性质。采用UCI标准库数据集进行实验分析,对比两种模型的实验结果,验证了改进邻域多粒度粗糙集模型在近似逼近方面的优越性。

关键词: 粗糙集, 邻域关系, 多粒度, 距离函数, 兰氏距离, 邻域半径

Abstract: At present, the neighborhood multi-granulation rough set widely adopting the Minkowski function is limited to some extend. In this paper, Lance distance is cited as the distance function, and reconstructs the neighborhood radius method. Then, an improved neighborhood rough multi-granulation rough set model is defined. In addition, some related properties are proven. The results of two models are compared through the experimental analysis on the UCI standard data sets, and also the superiority of the improved neighborhood multi-granulation rough set model is verified in terms of approximation approach.

Key words: rough set, neighborhood relation, multi-granulation, distance function, Lance distance, neighborhood radius