一种Gray区域表征的关键条件属性集逆向求解.pdfVIP

一种’G+ray区域表征的关键条件属性集逆向求解 程玉胜1·2 张佑生1 胡学钢1 (1合肥工业大学计算机与信息学院．安徽。合肥，230009) ‘ (2安庆师范学院计算机系，安徽，安庆．246011) 摘要：本文利用粗集理论定义了两个等价关系下的知识相对划分，它是后继元知识的生成和近似约筒的关键；另外建立了 信息损失率，信息保存率与Gray区域之间的联系，提出了／*1Gray区域表征的一种逆向求解关键条件属性集的盯乒-i匠似约简 方法，为在大数据库中较快的发现近似约简提供了可能．并通过一个例子，对本文提出的方法进行了验证． 关键词：粗集；近似约筒：Apriori算法 AConversel for ConditionAttributeSet ApproachComputingKey BasedonTokenof GrayRegion 1 2 l I Hu Yusheng．ChengYoushengZhangXuegang (ISchool ofComputer UniversityofTechnology，Hefei·230009) Scienc．e·Hefei (2DepofComputer，AnqingTeachersCollege，．Anqing，．246011) Abstract：Inthis roomisderivedtwo classesbasedon setandusedfor of paper,theapproximate by equivalence rough generalization SuccessorMeta and reduction．Ourfor reductionisto theinformation knowledgeapproximate approachapproximate computer for condition ratio／informationloss relatesto converse preserving ratio，denotedIPR／ILR，whichGrayregion．A computingkey by attributesetbasedontokenof is whichnamedo踮-reduction．ItforUStoderivethe condition forward is Grayregionput possible key attributesetfromdatabases isillustratedinthe and thatthe is effective． very efficiently．Anexample paperproves approach Sets Keywords：RoughTheory：ApproximateReduction：AprioriAlgorithm 1引言