【作者】 张铮； 喻道远；

【Author】 Zhang Zheng~1 Yu Daoyuan~2 (1 School of Mechanical Engineering,Hubei University of Technology,Wuhan 430068,China; 2 School of Mechanical Science & Engineering,Huazhong University of Science & Technology,Wuhan 430074,China)

【机构】 湖北工业大学机械工程学院； 华中科技大学机械科学与工程学院；

【摘要】 本文以可分辨矩阵为基础,引出了次核、再生分辨矩阵等新概念,进而提出了基于最大分辨度的启发式属性约简算法, 实现了求取决策表最小属性约简集一种有效方法。受认识规律的启发,将不完备信息决策系统分为完备层和非完备层,用最大分辨度的启发式属性约简算法先对完备层进行约简,然后将不完备属性根据完好度大小逐层分步约简并提取规则,减小因系统完备化使信息失真而带来的不确定性。实例表明,该算法不仅适合于完备决策信息系统,同时也适用于不完备决策信息系统的属性约简与规则获取。更多还原

【Abstract】 The concepts of secondary core and regenerated discernibility matrix are given based on the dis- cernibility matrix,and then the heuristic attribute reduction algorithm based on maximum discernibility de- gree is put forward,which reaches an effective way of gaining minimum attribute reduction of decision ta- ble.According to thinking of the hierarchical reduction approach,the hierarchical reduction approach based on the significance of attribute is put forward.The approach,which does not lead the information losing of decision system,satisfies the fast reduction requirements of dyhamical system,and has better practicability. The incomplete information system,enlightened on the recognition laws of human being,is divided into complete attribute level and incomplete attribute level.Firstly,complete attribute level can be reduced by the heuristic attribute reduction algorithm based on maximum discernibility degree.Secondly,the incom- plete attribute can be reduced according to its intactness degree one by one.Lastly,the rules can be extrac- ted rightly.The reduction approach can decrease the uncertainty because of completeness process.The practical results show this approach both are fit for the attribute reduce or rules acquisition of complete and incomplete information system.更多还原