Introduction to Machine Learning-stanford

《Introduction to Machine Learning-stanford》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、INTRODUCTIONTOMACHINELEARNINGANEARLYDRAFTOFAPROPOSEDTEXTBOOKNilsJ.NilssonRoboticsLaboratoryDepartmentofComputerScienceStanfordUniversityStanford,CA94305e-mail:nilsson@cs.stanford.eduNovember3,1998Copyrightc2005NilsJ.NilssonThismaterialmaynotbecopied,reproduced,ordistributedwi

2、thoutthewrittenpermissionofthecopyrightholder.iiContents1Preliminaries11.1Introduction..............................11.1.1WhatisMachineLearning?.................11.1.2WellspringsofMachineLearning..............31.1.3VarietiesofMachineLearning................41.2LearningInput-O

3、utputFunctions..................51.2.1TypesofLearning......................51.2.2InputVectors.........................71.2.3Outputs............................81.2.4TrainingRegimes.......................81.2.5Noise.............................91.2.6PerformanceEvaluation......

4、.............91.3LearningRequiresBias........................91.4SampleApplications.........................111.5Sources................................131.6BibliographicalandHistoricalRemarks..............132BooleanFunctions152.1Representation............................152.

5、1.1BooleanAlgebra.......................152.1.2DiagrammaticRepresentations...............162.2ClassesofBooleanFunctions....................172.2.1TermsandClauses......................172.2.2DNFFunctions........................182.2.3CNFFunctions........................212.2.4

6、DecisionLists.........................222.2.5SymmetricandVotingFunctions..............232.2.6LinearlySeparableFunctions................232.3Summary...............................242.4BibliographicalandHistoricalRemarks..............25iii3UsingVersionSpacesforLearning273.1Vers

7、ionSpacesandMistakeBounds................273.2VersionGraphs............................293.3LearningasSearchofaVersionSpace...............323.4TheCandidateEliminationMethod................323.5BibliographicalandHistoricalRemarks..............344NeuralNetworks354.1ThresholdLog

8、icUnits........................354.1.1DenitionsandGeometry.........