计算机工程与应用 ›› 2011, Vol. 47 ›› Issue (28): 48-51.

• 研究、探讨 • 上一篇    下一篇

混合值不完备决策信息系统的粗糙分类方法

黄恒秋,曾 玲   

  1. 桂林电子科技大学 数学与计算科学学院,广西 桂林 541004
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2011-10-01 发布日期:2011-10-01

Rough classification method in incomplete decision information system with hybrid value

HUANG Hengqiu,ZENG Ling   

  1. School of Mathematics and Computing Science,Guilin University of Electronic Technology,Guilin,Guangxi 541004,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2011-10-01 Published:2011-10-01

摘要: 针对混合值不完备决策信息系统,提出一种将邻域联系度粗糙集与贝叶斯理论相结合的分类方法。定义了一种新的属性辨识矩阵——同异反辨识矩阵,给出了基于同异反辨识矩阵的t分配约简算法,以及对约简后的决策信息系统建立基于邻域联系度粗糙集的最小错误率贝叶斯决策准则,用于对含有混合属性值以及不完备数据的对象进行分类。实验表明所提出的方法是客观有效的。

关键词: 粗糙集, 邻域联系度, 同异反辨识矩阵, 分配约简, 贝叶斯理论

Abstract: A classification method is presented based on the combination of neighborhood connection degree rough set and Bayesian theory in incomplete decision information system with hybrid value.A new attribute discernibility matrix—identical-
discrepancy-contrary discernibility matrix is defined,t-assignment reduction algorithm based on identical-discrepancy-contrary discernibility matrix is proposed.In addition,a Bayesian decision criterion of minimum error rate based on neighborhood connection degree rough set is established to classify the object with hybrid attribute value and incomplete data in reduction decision information system.Experiments show that new method is objective and feasible.

Key words: rough set, neighborhood connection degree, identical-discrepancy-contrary discernibility matrix, assignment reduction, Bayesian theory