基于新的条件熵的决策表约简方法  

Reduction method of decision table based on new conditional entropy

作　　者：徐久成 孙林[1] 马媛媛[1] 

机构地区：[1]河南师范大学计算机与信息技术学院,河南新乡453007

出　　处：《计算机工程与设计》2008年第9期2313-2316,共4页Computer Engineering and Design

基　　金：河南省自然科学基金项目(0511011500);河南省高校新世纪优秀人才支持计划基金项目(2006HANCET-19)

摘　　要：分析了在知识约简过程中现有条件熵的不足,在一致和不一致对象分开的基础上,定义了一种新的条件熵概念,以弥补现有信息熵的不足,在此基础上给出了以不等式为条件的约简判定定理;然后以条件属性子集的条件熵来度量其对决策分类的重要性,提出了一种新的知识约简启发式方法。应用实例分析的结果表明,基于新的条件熵的属性重要性是一种更准确、更有效的启发式信息,该方法时间复杂度较低,有助于搜索最小或次优知识约简。In decision table, the disadvantages of the current conditional entropy are analyzed deeply. To eliminate the limitations, a new conditional entropy is defined with separating consistent objects form inconsistent objects, and the judgment theorem with respect to knowledge reduction is obtained from inequality. Condition attributes are considered to estimate the significance for decision classes, and a heuristic algorithm is proposed. Theoretical analyses show that the proposed heuristic information is better and more efficient than the others, and its time complexity is relatively less. Experimental results prove the validity of this reduction method in searching the minimal or optimal reduction. So it enlarges the application area of rough set theory .

关 键 词：粗糙集 知识约简 决策表 条件熵 信息熵 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程] TP301.6[自动化与计算机技术—控制科学与工程]