预览加载中,请您耐心等待几秒...
1/10
2/10
3/10
4/10
5/10
6/10
7/10
8/10
9/10
10/10

亲,该文档总共31页,到这已经超出免费预览范围,如果喜欢就直接下载吧~

如果您无法下载资料,请参考说明:

1、部分资料下载需要金币,请确保您的账户上有足够的金币

2、已购买过的文档,再次下载不重复扣费

3、资料包下载后请先用软件解压,在使用对应软件打开

基于数据挖掘的企业失信识别模型研究数业161203105328·2020业信的业基于数据挖掘的企业失信识别模型研究的研究的的的数据2020业基于数据挖掘的企业失信识别模型研究企业失信数据的数据挖掘企业失信Python的pandas数据的失失30%的19企业失信的12R的DMwR基于knn数据的失数据的企业型企业4企业失信数据模型模型模型企业失信识别模型ROC模型的别90.9%92%92.57%模型的AUC0.510.61模型的MLP模型MLP模型的92%AUC0.89.企业失信数据挖掘ROC2020业ResearchonthemodelofcorporatedishonestyrecognitionbasedonDataMiningAbstractHowtodetectwhetherenterprisesbreakthelawornothasbecomeaproblemintheeraofbigdata.Thispaperwillusedataminingmethodtopredictwhethertheenterpriseisdishonest.Firstofall,weusePython'spandaspackagetomakestatisticsonthemissingvaluesofthedata,andremovetheindicatorswithamissingrateofmorethan30%,andthenremovetheremaining19indicatorsthathavenoimpactontheenterprise'sdishonesty.Atlast,weleave12indicators.Then,basedontheprincipleofKNNalgorithm,thedmwrpackageofRisusedtofillthemissingdata.Secondly,thispapermakesadatavisualiza-tionanalysisonthefourindicatorsofdata,namely,enterprisetype,registrationauthority,enterprisestatusandjurisdictionauthority.Finally,thedecisiontreemodel,randomforestmodelandgradientpromotiondecisiontreemodelareselectedtoestablishtheenterprisedishonestyidentificationmodel.Themodelisevaluatedbytheaccuracyrate,recallrate,confusionmatrixandROCcurve.Finally,thepredictionaccuracyratesofdecisiontree,randomforestandgradientpromotiondecisiontreeare90.9%,92%and92.57%.However,theAUCvalueofthethreedecisiontreefamilymodelsisintherangeof0.51and0.61,andthepredictionaccuracyofthemodelsisnotstableenough.Finally,wetryMLPmodel,and