資源描述：
《Advances in Large Margin Classifiers》由會(huì)員上傳分享，免費(fèi)在線閱讀，更多相關(guān)內(nèi)容在學(xué)術(shù)論文-天天文庫(kù)。
1、OFRELATEDINTERESTAdvancesinKernelMethodsSupportVectorLearningADVANCESINLARGEMARGINCLASSIFIERSeditedbyBernhardSch?lkopf,ChristopherJ.C.Burges,andAlexanderJ.SmolaADVANCESINTheSupportVectorMachineisapowerfulnewlearningalgorithmforsolvingaADVANCESINLARGEMARGINvarietyoflearningan
2、dfunctionestimationproblems,suchaspatternrecognition,regressionestimation,andoperatorinversion.CLASSIFIERSTheimpetusforthiscollectionwasaworkshoponSupportSMOLALARGEMARGINVectorMachinesheldatthe1997NIPSconference.TheeditedbyAlexanderJ.Smola,PeterJ.Bartlett,contributors,bothun
3、iversityresearchersandengineersBernhardSch?lkopf,andDaleSchuurmansdevelopingapplicationsforthecorporateworld,.CLASSIFIERSTheconceptoflargemarginsisaunifyingprincipleformaWho’sWhoofthisexcitingnewarea.BARTLETTfortheanalysisofmanydifferentapproachestotheclassificationofdatafro
4、mexamples,includingboost-ing,mathematicalprogramming,neuralnetworks,andsupportvectormachines.Thefactthatitisthemargin,orconfidencelevel,ofaclassification—that.is,ascaleparameter—ratherthanarawtraininger-SCH?LKOPFrorthatmattershasbecomeakeytoolfordealingwithclassifiers.Thisbo
5、okshowshowthisideaap-pliestoboththetheoreticalanalysisandthedesignofalgorithms.Thebookprovidesanoverviewofrecentdevelop-.mentsinlargemarginclassifiers,examinesconnec-SCHUURMANStionswithothermethods(e.g.,Bayesianinference),andidentifiesstrengthsandweaknessesofthemethod,aswell
6、asdirectionsforfutureresearch.AmongthecontributorsareManfredOpper,PulsedNeuralNetworksEDITEDBYVladimirVapnik,andGraceWahba.editedbyWolfgangMaassandChristopherM.BishopMostpracticalapplicationsofartificialneuralnetworksarebasedonaALEXANDERJ.SMOLAAlexanderJ.Smolaisaresearcherin
7、theDepart-computationalmodelinvolvingthepropagationofcontinuousvari-ablesfromoneprocessingunittothenext.Inrecentyears,dataEDITORSPETERL.BARTLETTmentofEngineeringandRSISE,AustralianNationalUniversity.PeterL.BartlettisSeniorFellow,Com-fromneurobiologicalexperimentshavemadeitin
8、creasinglyBERNHARDSCH?LKOPFputerSciencesLaboratory,AustralianNationalUni-cl
