欢迎来到天天文库
浏览记录
ID:36522816
大小:1.44 MB
页数:51页
时间:2019-05-11
《粗糙集理论在数据库中的应用研究》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库。
1、山东大学硕士学位论文粗糙集理论在数据库中的应用研究姓名:霍雯申请学位级别:硕士专业:计算机软件与理论指导教师:张世栋20090405⋯东大学硕士学位论文3、在l、2的基础上设计了一种基于划分加细和一致度的启发式知识约简算法,此算法的时间复杂度为o(Icl2IU
2、),其中ICJ为条件属性个数,lUI为论域U中个体的数目,低于现有的经典知识约简算法,而且计算量较小。4、率先提出了从粗糙关系数据结构、粗糙关系操作、粗糙关系完整性约束、粗糙关系规范化这四个方面,建立一个完整的粗糙关系数据库模型;5、提出了粗糙关系完整性约束,以完善粗糙关系数据库模型对不完全信息的处
3、理能力:6、提出了粗糙关系规范化理论及相应的规范化算法,以解决在粗糙关系数据库逻辑设计中如何构造一个好的数据库模式问题。关键词:粗糙集;数据库中的知识发现;知识约简:决策表;粗糙关系数据库。Il山东大学硕士学位论文ABSTRACTRoughSetTheoryisanewmathematicaltooltodeal、Ⅳimvagueanduncertainproblems,whichhasbeenwidelyappliedinmachineleaming,decisionanalysis,knowledgediscovery,expertsystem,dec
4、isionsupportsystem,patternrecognition,fuzzycontrolorotherfieldsAtpresenttheresearchOilapplicationofRoughSetTheoryindatabasefocusesontwoaspects:OneaspectisKDD(KnowledgeDiscoveryinDatabase),theotheraspectisRRDM(RoughRelationalDatabaseModel)Knowledgereduction,whichisalsonamedattribut
5、ereduction,isoneofthemainproblemsinKDDthatRoughSetTheorydeals、Ⅳim.ThetimecomplexityofpresentknowledgealgorithmsthatbasedondiscerniblematrixorbasedondiscriminationfunctionisgenerallyO(IAl2IU]2),whereiSthenumberofattributesandIulisthenumberofelementsindiscoursedomainU.Whenthedataamo
6、utisverylarge,thefeasibilityofthesealgorithmswillfacechallenge.AndthelowefficiencyofthesealgorithmshasalsolimitedthewideapplicationofRoughSetTheory.SoithasgreatsignificanceinseekingthealgorithmwithhighefficiencyandgoodfeasibilityRoughRelationalDatabaseModel(RRDM)isacompositeproduc
7、tionofRoughSetTheoryandclassicalRelationalDatabaseModel.NowinthestudyonRRDM,thedomesticandforeignscholarsstudiedsomeaspectsseparatelywithoutcombiningthemtogether,andsomeconceptsofRRDMisnotdefinednormally.SoifanintegratedandnormaldescriptionofRRDMisconstructedfromfouraspects:therou
8、ghrelationaldatastructure,theroughrelationaldatabaseoperations,theroughrelationalintegrityconstraintandtheroughrelationalnormalization,itwillcertainlylayacompletetheoreticalfoundationoftheimplementationandthewideapplicationofRRDM.Theresearchworkofthispaperfocusesontwoaspects:oneas
9、pectistoseekknowledgereductionalg
此文档下载收益归作者所有