资源描述:
《Introduction to semi-supervised learning by WISC xiaojin ZHU》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库。
1、IntroductiontoSemi-SupervisedLearningSynthesisLecturesonArtificialIntelligenceandMachineLearningEditorsRonaldJ.Brachman,Yahoo!ResearchThomasDietterich,OregonStateUniversityIntroductiontoSemi-SupervisedLearningXiaojinZhuandAndrewB.Goldberg2009ActionProgrammingLanguagesMichaelThielscher2008Represent
2、ationDiscoveryusingHarmonicAnalysisSridharMahadevan2008EssentialsofGameTheory:AConciseMultidisciplinaryIntroductionKevinLeyton-Brown,YoavShoham2008AConciseIntroductiontoMultiagentSystemsandDistributedArtificialIntelligenceNikosVlassis2007IntelligentAutonomousRobotics:ARobotSoccerCaseStudyPeterSton
3、e2007Copyright©2009byMorgan&ClaypoolAllrightsreserved.Nopartofthispublicationmaybereproduced,storedinaretrievalsystem,ortransmittedinanyformorbyanymeanselectronic,mechanical,photocopy,recording,oranyotherexceptforbriefquotationsinprintedreviews,withoutthepriorpermissionofthepublisher.Introduction
4、toSemi-SupervisedLearningXiaojinZhuandAndrewB.Goldbergwww.morganclaypool.comISBN:9781598295474paperbackISBN:9781598295481ebookDOI10.2200/S00196ED1V01Y200906AIM006APublicationintheMorgan&ClaypoolPublishersseriesSYNTHESISLECTURESONARTIFICIALINTELLIGENCEANDMACHINELEARNINGLecture#6SeriesEditors:Ronal
5、dJ.Brachman,Yahoo!ResearchThomasDietterich,OregonStateUniversitySeriesISSNSynthesisLecturesonArtificialIntelligenceandMachineLearningPrint1939-4608Electronic1939-4616IntroductiontoSemi-SupervisedLearningXiaojinZhuandAndrewB.GoldbergUniversityofWisconsin,MadisonSYNTHESISLECTURESONARTIFICIALINTELLIG
6、ENCEANDMACHINELEARNING#6MMorgan&cLaypoolpublishers&CABSTRACTSemi-supervisedlearningisalearningparadigmconcernedwiththestudyofhowcomputersandnaturalsystemssuchashumanslearninthepresenceofbothlabeledandunlabeleddata.Traditionally,learninghasbeenstudiedeitherintheunsupervisedparadigm(e.g.,clustering
7、,outlierdetection)whereallthedataisunlabeled,orinthesupervisedparadigm(e.g.,classification,regression)whereallthedataislabeled.Thegoalofsemi-supervisedlearningistounderstandhowcombininglabeledandunlabeleddatamaychangeth