一种新的信息熵属性约简算法  

Novel attribution reduction algorithm of information entropy

作　　者：舒文豪[1] 徐章艳[1,2] 杨炳儒[2] 钱文彬[1] 

机构地区：[1]广西师范大学计算机科学与信息工程学院,广西桂林541004 [2]北京科技大学信息工程学院,北京100083

出　　处：《计算机工程与应用》2009年第32期108-110,113,共4页Computer Engineering and Applications

基　　金：国家自然科学基金No.60463003;广西教育厅科研基金项目No.200807MS015;广西研究生教育创新计划项目~~

摘　　要：给出一个区分对象对的属性约简定义,同时证明该属性约简的定义与基于信息熵的属性约简的定义是等价的。为求出区分对象对集,首先给出了一个快速求简化决策表的算法,其时间复杂度为O(|C||U)|。然后在简化决策表的基础上,设计了基于区分对象对集的信息熵属性约简算法,其时间复杂度和空间复杂度分别为O(|C||U)|+O(|C||U/C|2)和O(|U/C|2)+O(|U|),最后用一个实例说明了新算法的高效性。Firstly,the definition of attribution reduction of discernibility object pair set is provided.At the same time,it is proved that the above definition of attribution reduction is equal to the definition of attribution reduction based on information entropy.In order to compute discernibility object pair set,a quick algorithm for simplified decision table is provided,whose time complexity is 0 （｜C｜｜U｜）.On this condition,an efficient attribution reduction algorithm based on information entropy of discernibility object pair set is designed,whose time complexity and space complexity of the new algorithm are O（｜C｜｜U｜）＋O（｜C｜｜U/CI^2） and O（｜U/CI^2）＋ O（｜U｜）respectively.Finally,an example is used to illustrate the efficiency of the new algorithm.

关 键 词：粗糙集 信息熵 区分对象对集 属性约简 算法复杂度 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]