基于测试代价敏感的不完备决策系统属性约简算法  

ATTRIBUTE REDUCTION ALGORITHM OF INCOMPLETE DECISION SYSTEM BASED ON TEST COST SENSITIVITY

作　　者：谢小军[1,2] 徐章艳[1,2] 乔丽娟[1,2] 朱金虎[1,2] 

机构地区：[1]广西多源信息挖掘与安全重点实验室,广西桂林541004 [2]广西师范大学计算机科学与信息工程学院,广西桂林541004

出　　处：《计算机应用与软件》2016年第9期264-268,295,共6页Computer Applications and Software

基　　金：国家自然科学基金项目(61262004;61363034;60963008);广西自然科学基金项目(2011GXNSFA018163);八桂学者专项基金

摘　　要：提出不完备决策系统测试代价敏感属性约简问题,给出不一致对象集定义以及求解不一致对象集的算法。根据不一致对象的性质改进属性重要性定义,考虑测试代价因素以及不一致对象个数的改变量给出一个新的属性重要性的定义和属性重要性中权重的设置方法,并给出属性重要性的计算算法。在此基础上,给出一个时间复杂度为O(k|C|2|U|)和空间复杂度为O(|U|)的启发式属性约简算法,并通过理论分析、实例分析和实验分析说明该算法准确性和可行性。We introduced the problem of test-cost-sensitive attribute reduction in incomplete decision system, and suggested the definition of inconsistent object set and an algorithm for computing the inconsistent object set. According to the nature of inconsistent object set we improved the definition of attribute significance. Considering the test cost factors ani：l the varied amount of the number of inconsistent objects we presented a new definition of attribute significance and the weight setting method of it. And then we gave the calculation algorithm of attribute significance. Based on these conditions, we proposed a heuristic attribute reduction algorithm with the time complexity O（k｜C｜^2｜U｜） and the space complexity O（｜U｜）. Through theoretical analysis, example analysis and experiment analysis we explained the accuracy and feasibility of the reduction algorithm.

关 键 词：测试代价敏感 不完备决策系统 属性重要性 属性约简 不一致对象 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]