Looking inside ensembles of negatively correlated Self-Organizing Maps. [Elektronische Ressource] / Alexandra Scherbart. Technische Fakultät - AG Neuroinformatik

Publié par

desBielefeldvUniverersittyAlexandraFtacultsyoofGradesTuaryecthnologylBioDdataeMiningt&zurAppliedademiscNveuroinformaticsherbartGroup2011Losio?kinBielefegdInsideorgelegteEnsemblesiofsnegativelyrcoarrelatediSelf-OrganizingnMapsErlangungDerakThenecDr.-Ing.hnisconhenScFJanakult?t9,derUniverSupNattkervisor:W.JunProf.empDr.TimypingList.of.FiguresHandlingvii.List.of.T.ables.ix.Glossavry.xi.1.Intro.duction.1.1.1erformancePComparisoneak.In.teeighn.si15t.yEvPrediction......25.......P...2.3.5.4.............b.the.......V.......on..1.1.2.Bias2.3.4and.V.ariance.and.the.T.rade-O..2.3.5.1...........2.3.5.3.NG.eptides.....of.Discussion.......33......3.1.3InWhey.applyeV.ector-Quan.tizLeastat.i.on.Based-SuppSelf-OrganizingMacMaps?..........20.uat4.1.4.Wh.y.Ens.e.m.ble.Learning?in.............Results.................eptide...........2.3.5.2.In......6.1.5PredictionPSOMoten.tialofof.SOM.Ensem.ble.Learning29.P.eature.31.................Conclusions...............2.47P1.6yOutlineA.eature...2.4.1.eature.ance.........2.4.1.1.uares.........38...2.3.2Contents.ort.ector.hine.......................2.3.3.al.i..9.1.7.Publications........................23.Issues.Data.........................2.3.5..............10.2.P.eak.Intensit.y.Prediction.in.a.Single26LeaPrnerPrototS.etup.11.2.1.MS.data..........26.Predicting.eaks'.tensities...............27.Comparison.P.of.to...28.Subsampling.P...................2.3.5.5.Prediction.erformance13F2.2SetsBenc.hmark2.3.6Datasets................................2.3.7.....
Publié le : samedi 1 janvier 2011
Lecture(s) : 16
Tags :
Source : D-NB.INFO/1015206379/34
Nombre de pages : 150
Voir plus Voir moins

desBielefeldvUniverersittyAlexandraFtacultsyoofGradesTuaryecthnologylBioDdataeMiningt&zurAppliedademiscNveuroinformaticsherbartGroup2011Losio?kinBielefegdInsideorgelegteEnsemblesiofsnegativelyrcoarrelatediSelf-OrganizingnMapsErlangungDerakThenecDr.-Ing.hnisconhenScFJanakult?t9,derUniverSupNattkervisor:W.JunProf.empDr.Tim
ypingList.of.FiguresHandlingvii.List.of.T.ables.ix.Glossavry.xi.1.Intro.duction.1.1.1erformancePComparisoneak.In.teeighn.si15t.yEvPrediction......25.......P...2.3.5.4.............b.the.......V.......on..1.1.2.Bias2.3.4and.V.ariance.and.the.T.rade-O..2.3.5.1...........2.3.5.3.NG.eptides.....of.Discussion.......33......3.1.3InWhey.applyeV.ector-Quan.tizLeastat.i.on.Based-SuppSelf-OrganizingMacMaps?..........20.uat4.1.4.Wh.y.Ens.e.m.ble.Learning?in.............Results.................eptide...........2.3.5.2.In......6.1.5PredictionPSOMoten.tialofof.SOM.Ensem.ble.Learning29.P.eature.31.................Conclusions...............2.47P1.6yOutlineA.eature...2.4.1.eature.ance.........2.4.1.1.uares.........38...2.3.2Contents.ort.ector.hine.......................2.3.3.al.i..9.1.7.Publications........................23.Issues.Data.........................2.3.5..............10.2.P.eak.Intensit.y.Prediction.in.a.Single26LeaPrnerPrototS.etup.11.2.1.MS.data..........26.Predicting.eaks'.tensities...............27.Comparison.P.of.to...28.Subsampling.P...................2.3.5.5.Prediction.erformance13F2.2SetsBenc.hmark2.3.6Datasets................................2.3.7..........................13.2.3.T34oImprowedardseakPtensiteaPredictionkyIndaptivtensitFyWPreditingction.with35MacAssessinghineFLearningRmetholevds.......15.2.3.1.Lo.cal.Linear.Map37(LLM)Linear-SqV.Q-based.approac.h..............iii..Con.ten.tsDiverse2.4.1.2.P.artialonLe.as.t.Squares....Quan.........Strengthening.......P.ork.............Algorithm385.32.4.1.3.Random.F.orests..tor.....Metho.....e.....53...4.2...v.............5.aluation38.2.4.1.4.Bagged.T.rees....Data.............y...51.......Random.....Random............39y2.4.2.Ev.a57luationble.....Arc.....A.Predictors.....tra-SOM.....4.5.Learning...Conclusion.........cc.rs.........66......39.2.4.3ofResults......69.........Size...............50.Eect.erse.......Bagging.............51.pac....40.2.4.4.W52eighoreststed.F.eature.Space....3.3.2.Learning.........52.P.unctions.........LERRANCO.Related.SOM...........osed41.2.4.5.W.eigh.ted.and58FilteredandFEnsemeature.Space......61.of.ersit...........W.Ensem.NCL.....6441.2.5.Discussion..............as.rate.emble.5.1...................T.................op.w..........42.2.6.ConclusionsT.................69.Random...........70...................3.3.1.the.b.Div.Predic.s......43.2.7.Con3.3.1.1tribution.to.Op.enMS.-.An.Op.en-Source.F.ramew.ork.for.MS....3.3.1.2.Subs.e.d44.3.Ensemble.Lea.rning.45.3.1.Reasons.for3.3.1.3theFSuccess.of.Ensem.bles................52.Negativ.Correlation....................463.3.33.1.1arameterizingWenalteakFLearners..................4.Architecture.4.1.W.on.Ensem.Learning.................57.Prop.LERRANCO.hitecture..47.3.1.2.Unstable.Learners............4.3.ccurate.Di.erse.ble...................4.4.tication.In.Div.y....47.3.1.3.History.of.Ensem.bles....63.Related.ork.SOM.ble.with.............4.6......................47.3.2.Assessing.Ensem.ble.Error..64.SOMs.A.u.and.Ens.Predicto.65.Ev..................................48.3.2.15.2Bias-Vrainingaria.n.c.e-Co.v.ariance.Decomp.osition................66.T.ology.Net.orks..48.3.2.2.Am.biguit.y.Deco.mp.osition............5.4.raining..............................49.3.3.Assessing5.4.1DivofersitSubspacesy......................iv.
9.4ten109ts.5.5.Initial.Conditionstext.......................Exp.Predic.......Random.eature.Robustness.Con...Conclusion.............Ensemble....71.5.5.17.3Grid.Size......8.1.......osteriori.on.......116.........Ho...Wh...9.3.........v...of.Aggregation....71.5.5.2AggregationGaussian.Neigh.b.orho.o.d.Width............F.ance.Learning...Subspace.....8.3.on...eature.p.R..73.5.6.Discussioneatures.......................9.bles.......ensem.................................7.otheses75995.7LoConclusion............99.Ensem.s.............................7.4.................104.Relevance.eature77the6EnsemNegatively.Co.rrelated106SeighOMdEnsembles.79.6.1.LERRANCO.Ev.aluationthe.ance.a.....the.ance.Mappings.112.F.lev...........113.of.Visualization.........Discussion........79.6.2.Results....119...................121.SOM.............121.SO.succeed.............ok................80.6.3.In.ter-SOM.Div.ersit.y........Bibliography.............96.Aggregation.Hyp.to.Prediction.7.1.of.cal.erts....................85.6.4.In7.2tra-SOMofDivbleersittory....................100.Discussion...................................10385Conclusion6.5.Dynamics.of.Bo.osted.Negativ.ely.Correlated.SOMs..................8.eature.105.F87Relev6.5.1inDynamicsConinofTimeble...........8.2.W.ted.Metho.....................106.Assessing.F.Relev.based.OOB.p87.6.5.2.Dynamics.in.Con8.4Assessing.F.Relev.based.Linear.a.osteriori.8.5.of.eature.e.ance.......................8.6.tribution.F.to....91.6.6.Supp.orting.Altered.P.enalt.y8.7F.u.nc.tions..............................8.8........91.6.7.Discussion......................119.Conclusions.9.1.w.ensem.succeed.........................9.2.y.M.bles......93.6.8.Complexit.y.and.Computation.Time......122.Outlo................................95.6.9.Conclusion123.Summary...................................124.127.."

K
K nl
= 0
K nl
K
0


and.,.Scatterplots.y.s...56...n....16.2.26.6Neural.Gasbsim.ulation.sequence......83.....88.c...prop.the.MSE.M.v.5.3.des.T.....,.......y17ulation2.3.Grovwing.Neural.Gas:.Insertionestof89atimenew.no92deo.ble.hematic.space...v.ble...MSE.randomly...when.b.....arying18.2.4.Gro.wingforNeuralestGas.sim6.1ulation.sequence........hmark...est.......MSE.Self-Organizing..............19,2.5.Scforhematicerrepresen.tationdynamicsof.LLM.training..Sc.erview.LERRANCO.hitecture.4.2.tation.yp.div...63.est.n.of.b.....7020est2.6sizeSVMd-.Separating.hTyparyingerplanensoffor.t.w.o5.4classeswhen.,.............5.5.,.MSE.MSE.in......21.2.7.SVR.-.Figures.-insensitiv.e......82.b.predicted.....MSE.v.sim...........84.est.Map.....6.4..................23.2.8Scatterplot-matricesPMSEarallelMSEcoinordinatesinplot.forMSEdatasetriedmanHeuro.rai.....sho.................4.1.hematic.v.of.osed.ensem27arc2.9.Scatterplots59ofScdatasetrepresenHeurof.h.othesis.and.ersit.......5.1.T.when.arying.um.er.ensem.mem.ers...........5.2.T.when.arying29of2.10selecteResultssubspacesof.predi.ction.accurac72yMSE-estSOMvvs.NGthe.um.er.no.............73.MSE.est.v.e.c30ue2.11.Scatterplots.of.su.bsampl.ed.p.eptides........74.Scatterplot-matrices.q.e.,.T.,.Map.DIV.ter.......76......32.2.12.Bet.w.een.p.eptide.correlations......................6.1.of.enc.datasets.b.LERRANCO.......6.2.T.when.arying.and34.2.13.Sc.hematic.represen.tation.of.A.daptiv.e......6.3.T.when.arying.,.2.1.F.eature.W.eigh.ting..............86....36.2.14.Lev.elplot.of.AA.index.F.eature.Relev.ance..................6.4.for.,.T.,.Map.DIV.ter.DIV.tra40.3.1.Upp6.5erMapbFounddataofhangingofvdeptendingningon.List.......90.Violinplots.wing.in.......................vii.bleListDierencesof.Fdictorsi8.2gurbasede.s.7.1.Strengthonof112Gaus.sia.n115Neigh.boforho.o.dRelev.....eature.....eature.....ensem.......ts...viii.....F.e.data100.7.2.MSE.TAssessedestancwhenlinearv.arying.smo114othnessassessedofanceoutput.function......Visualization.pr...........8.6101comp7.3fMSE.T.est.when.c.hanging.aggre.gation111toAssessedformeatureensemancblebasedpOOBre.dictor...........1038.38.1FScRelevhematicerepresenontationmappingsof.Random.W.eigh.ting.Subspace8.4MethoindF.Relev....107.8.2............8.5.of.ble.e.....................117.Visualization.main.onen.o.ensem.predictors.........118.. = 0

.2.1.DatasetforSurv.eyT...T.......102...des...for...w.......aggregation.hmark.....b.i.....6.1.datasets.....6.2.div.....p14.2.2.Prediction7.1capabilitiesdatasetsof.LLMestNGWSMfor.dataset.Heur....5.2.of.mi.um...........T.enc28.2.3.Prediction.test.error.for.datasetsbAAindexinandyablesHe.ur....6.3.for.y...........T.enc31v2.4.Prediction.accuracyMSEofbregressionwithmo.dels.o.fixLLM-t.yp.e......68.Num.er.no.with.n.m.MSE41est2.5.Statistics.on.feature.rel.ev.ance.for.dataset68AAindeMSExest.b.hmark....................43.4.181DivCorrelationersitetyeenbter-SOMoersitostingandfactors..............85.MSE.est.altered.enalt.functions...................93.MSE.est.b.hmark6when2arying5.1.MSE.T.est.for8.1bTencforhmarkencdatasets,datasetsTRof.List.............108..

Soyez le premier à déposer un commentaire !

17/1000 caractères maximum.