搜索
(,100871):,,,,,.:;;;:F830.5:A:1007-9807(2003)05-0046-070,.,,19979,8,.,,,[1].,,,,[2].,.,,21[3].,,,,[4].,(),.,(01),,,,.,,,,,,65200310JOURNALOFMANAGEMENTSCIENCESINCHINAVol.6No.5Oct.,2003:2001-10-24;:2003-06-25.:(02BJY126).:(1974),,,.,.12030,,,.,[5],,(),,.(),,,.Fisher[6],[7]LogisticFactor2logisticK[8][9].,,,,,[10],.,.,(AI).,,,,,[11,12].,,.,[13],,,,.,,;,.GreenSmith[14],,[15].,,(),,,,.,,,.22.1,,,.,,745:,,,.(),.,.,.(),,,,,,,,,.,,,,,,.,,,.,,,,().,01,,.,,.,,.,.2.2,,.(),,,,.:Dki=BkiSki(1):Dkiki;Bkiki;Skiki;mi=1Ski=Sk,Skk;ii.,:,,();,,01([0,1]),;,,;,,.,1.1{0,1}[0,1][0,1],,842003103Levenberg2Marquardt3.1,,,,.:.,;,;.,,.,.i,neti,fi,.,,.,,.3.2Levenberg2MarquardtBP,,Levenberg2Marquardt(LM).LM,.E=12p(p)2=122(2):pp,p.wold,wnew,wnew-wold,Taylor(wnew)=(wold)+Z(wnew-wold)(3),ZJacobian,(Z)pi=9p9wi(4)E=12(wold)+Z(wnew-wold)2(5)wnewE,wnew=wold-(ZTZ)-1ZT(wold)(6):T.E(2),HessianHik=92E9wi9wkp[9p9wi9p9wk+p92p9wi9wk](7)2,HessianH=ZTZ(8)(6)H-1,.,E=12(wold)+Z(wnew-wold)2+wnew-wold2(9)Ewnew,wnew=wold-(ZTZ+I)-1ZT(wold)(10),,(-1).E,E.44.1,(),.,,,//945:16,,,.160,636,76,560.,(,),200.,,SPSS.,x1,x2,,xk,xi,k=16.(Bartletttestofsphericity)KMO(Kaiser2Meyer2Olkinmeasureofsamplingadequacy),2.2KMOKaiser2Meyer2OlkinMeasureofSamplingAdequacy0.751BartlettsTestofSphericityApprox.Chi2Square14352.1188Df120Sig.0.0002,.,(1),4,(ScreeTestCriterion),74%,,,.,0.3,10%,0.3.,0.3.3123412341/0.9660.1950.052-0.0330.03220.9660.1950.052-0.0330.03230.9530.1920.022-0.038-0.00940.8430.1700.076-0.0290.06450.8190.1650.025-0.0450.10460.7550.152-0.037-0.014-0.22770.899-0.0320.2760.009-0.0538/0.891-0.0400.2730.014-0.13690.782-0.0380.2400.0040.249100.750-0.0240.2300.0070.23011-0.5660.3530.012-0.1740.0040.338120.8790.0480.0050.3250.037130.8430.0500.0050.3120.045140.8230.0090.0090.304-0.117150.6940.007-0.0220.257-0.065160.821-0.004-0.0280.0370.786052003103,,.4.2,50,100,.:..HechtNielsen,,,.,,,[16].,(1000,0.0010.0001,0.001,100.11e10),,(,,).4NodesMAPE/%NodesMAPE/%NodesMAPE/%NodesMAPE/%T8L819.52L8T820.11T1622.29L1623.24T9L917.58L9T916.48T1822.53L1821.78T10L1010.74L10T1011.53T2019.79L2019.05T11L1113.96L11T1114.14T2218.25L2220.21T12L1213.72L12T1216.76T2420.01L2419.35T13L1316.09L13T1317.14T2621.09L2621.87T14L1415.57L14T1414.35T2821.84L2821.26:Nodes;MAPE=1Neiyi,yi,ei,N,(SSE,MAE,MSE).T8L88sigmoid8sigmoid.,T16,16sigmoid.,:,,sigmoidsigmoid,1,4.Matlab5.3.5,,:(1),;(2),(sigmoidsigmoid);(3),,155:sigmoidsigmoid;(4)20,,.LM,1sigmoid(-,+)(-1,+1),2sigmoid(0,+1),,,(-,+),,,[0,1].6,,,,LM.,(),.,.,,().,,.,,.,,,,.:[1].[M].:,1997[2],.[M].:,1995[3]AltmanEI,CaoutteJB,NarayananP.Credit2riskmeasurementmanagement:Theironicchallengeinthenextdecade[J].Finan2cialAnalysisJournal,1998,(9):711[4]AltmanEI,SaundersA.Creditriskmeasurement:Developmentoverthelast20years[J].JournalofBankingandFinance,1998,(20):17211742[5],,.[J].,1999,13(1):58,30[6]FisherRA.Theuseofmultiplemeasurementsintaxonomicproblems[J].Ann.Eugenics,1936,7:179188[7]AltmanEL.Financialratios,discriminantanalysisandthepredictionofcorporatebankruptcy[J].JournalofFinance,1968,23(3):589609[8]TamKY,KiangM.Managerialapplicationsofneuralnetworks:Thecaseofbankfailurepredictions[J].ManagementScience,1992,38:927947[9]LundyM.Clusteranalysisincreditscoring[A].CreditScoringandCreditControl[M].NewYork:OxfordUniversityPress,1993.2536[10]GordyMB.Acomparativeanatomyofcreditriskmodels[J].JournalofBankingandFinance,2000,(24):119149[11]ThorsenPoddig.Bankruptcyprediction:Acomparisonwithdiscriminantanalysis[A].NeuralNetworksintheCapitalMarkets[M].editedbyApostolos2PaulRefenes,1992.311323(98)25200310[82],.[J].,1998,9(4):8892SurveyofmarketmicrostructuretheoryLIPing,ZENGYong,TANGXiao2woSchoolofManagement,UniversityofElectronicScienceandTechnologyofChina,Chengdu610054,ChinaAbstract:Inthepasttwentyyears,marketmicrostructurehasgrownsodramaticallythatitisnowarecognizedsub2fieldwithinfinance.Marketmicrostructureresearchisvaluableforilluminatingthebehaviorofpriceandmarkets.Thishasimmediateapplicationintheregulationanddesignofnewtradingmechanisms.Thispapersystematicallyreviewtheexistentresearchonmicrostructuredomesticandaboard.Thenintermofmarketmicrostructuretheory,weanalyzethemicrostructureoftheChinesestockmarket,andbringforwardsomeinstructiveproposalsconcerningthedevelopmentoftheChinesestockmarket.Keywords:marketmicrostructure;pricebehavior;tradingmechanism(52)[12],.[J].,1998,(9):2027[13]FrydmanH,AltmanEI,KaoDuen2Li.Introducingrecursivepartitioningforfinancialclassification:Thecaseoffinancialdistress[J].JournalofFinance,1985,40(1):269291[14]GreeneDP,SmithSF.AGeneticSystemforLearningModelsofConsumerChoice[C].Proceeding2ndInternationalConferenceonGA.Hillsdale,1987.217223[15],.[J].,2001,2:7379[16].[M].:,1999StudyoncreditriskassessingandforecastingmodelincommercialbankYULi2yongGuanghuaSchoolofManagement,PekingUniversity,Beijing100871,ChinaAbstract:Foralongtimecreditriskassessmenthasbeenregardedasonekindofclassificationissueofpatterni2dentification,whichcouldnotfullymeettherequirementsofcreditriskdecision2making.Inviewoftheconnotationofcreditrisk,itisputforwardthattheprobabilitythatcreditcapi