《数据挖掘开源工具weka简明教程》由会员分享,可在线阅读,更多相关《数据挖掘开源工具weka简明教程(173页珍藏版)》请在金锄头文库上搜索。
1、口WEKA:AMachineMachineLearningwithLearningToolkitWEKA口TheExplorer.ClassificationandRegression.ClusteringEibeank.AssociationRulesDepartmentlofGampaterScipnpes一亦A叮b踢s气】蹿叟。UniversitysfWaikatoyNewZealagd0|一“DAt4Visualization气TheExperimenter口TheKnowledgeFlowGUI日ConclusionsWEKA:thebird一JinchutouscoInWEKA:t
2、hesoftwareMachinelearning/dataminingsoftwarewritteninJava(distributedundertheGNUPublicLicense)口Usedforresearch,education,andapplications“Complements“DataMmmgbyWitten&Frank。Mainfeatures:4Comprehensivesetofdatapre-processingtools,learningalgorithmsandevaluationmethods*Graphicaluserinterfaces(incl.data
3、visualization)4Environmentforcomparinglearningalgorithms102112013UniversiyofWakatoWEKA:versionsThereareseveralversionsofWEKA:4WEKA3.0:“bookversion“compatiblewithdescriptionindataminingbook4WEKA3.2:“GUIversion“addsgraphicaluserinterfaces(bookversidnisc0mmand姻叮eonly)*WEKA3.3:“dsvelopmenitversiofrWithl
4、otsofimprovements口ThistalkisbasedonthelatestsnapshotofWEKA3.3(soontobeWEKA3.4)102112013UniversiyofWakatoWEKAonlydealswith“flat“filesrelationheartdisease-Simplifedattributeagenumericattributesexfemale,maleattributechest_pain_typetyp_angina,asymptnon_anginal,atyp_anginajattributecholesterolnumericatri
5、buteexercise_induced_anginanoyesattributeclasspiesenitnot.Fiesehti|data63,male:typ_angina,233,nonot_present67,male,asympt:286,yes,presentl扇latfilein67,male,asympt229,yes,present38,femalenon_anginal?,no,not_presentARFFf0rIna102112013UniversiyofWakato5WEKAonlydealswith“flat“files瞻_气【iohea鬣dis9a晕esimpe
6、dIl!lmericattribl工te0Tneeraenominalattributeattributechest_pain_typetyp_anginaasymptnon_anginal,atyp_anginajattributecholesterolnumericatributeexercise_induced_anginanoyesattributeclasspiesenitnotFiesehtidata63,male,typ_angina,233,no,not_present67,male,asympt;286,yes,present67,male,asympt229,yes,pre
7、sent38,femalenon_anginal?,no,not_present102112013UniversiyofWakatowecucieeer一“sss一WalcatoEvionmentiornowedosAnaysis(1999-2003UnvestyofWallatNewZealand102112013目_WetaculchooserWalcatoEnvionmentiorsszeEeRzzrKnowedgeAnalysis22*的1999-2003UnivessiyofyWallat102112013日_WetaculchooserWalcatoEnvionmentiorKno
8、wedgenabysis的1999-2003UnivesiyofWallato102112013Explorer:pre-processingthedataDatacanbeimportedfromafileinvariousformats:ARFF,CSV,C4.5,binaryDatacanalsobereadfromaURLorfromanSQLdatabase(usingJDBC)Pre-processingioglsjhWEKAarercallsd“filters“口WEKAcontainsfiltersfor:4Discretization,normalization,resampling,attributeselection,transformingandcombiningattributes,102112013UniversiyofWakalo10