商務(wù)統(tǒng)計全套配套課件[美]AmirD.AczelJayavelSounderpandian著譚英平等譯Chap011.ppt

《商務(wù)統(tǒng)計全套配套課件[美]AmirD.AczelJayavelSounderpandian著譚英平等譯Chap011.ppt》由會員上傳分享，免費在線閱讀，更多相關(guān)內(nèi)容在教育資源-天天文庫。

1、COMPLETEBUSINESSSTATISTICSbyAMIRD.ACZEL&JAYAVELSOUNDERPANDIAN6thedition.Chapter11MultipleRegressionUsingStatisticsThek-VariableMultipleRegressionModelTheFTestofaMultipleRegressionModelHowGoodistheRegressionTestsoftheSignificanceofIndividualRegressionParametersTestingth

2、eValidityoftheRegressionModelUsingtheMultipleRegressionModelforPredictionMultipleRegression(1)11QualitativeIndependentVariablesPolynomialRegressionNonlinearModelsandTransformationsMulticollinearityResidualAutocorrelationandtheDurbin-WatsonTestPartialFTestsandVariableSele

3、ctionMethodsMultipleRegressionUsingtheSolverTheMatrixApproachtoMultipleRegressionAnalysisMultipleRegression(2)11DeterminewhethermultipleregressionwouldbeapplicabletoagiveninstanceFormulateamultipleregressionmodelCarryoutamultipleregressionusingaspreadsheettemplateTestthe

4、validityofamultipleregressionbyanalyzingresidualsCarryouthypothesistestsabouttheregressioncoefficientsComputeapredictionintervalforthedependentvariableLEARNINGOBJECTIVES(1)11Afterstudyingthischapteryoushouldbeableto:UseindicatorvariablesinamultipleregressionCarryoutapoly

5、nomialregressionConductaDurbin-WatsontestforautocorrelationinresidualsConductapartialF-testDeterminewhichindependentvariablesaretobeincludedinamultipleregressionmodelSolvemultipleregressionproblemsusingtheSolvermacroLEARNINGOBJECTIVES(2)11Afterstudyingthischapteryoushoul

6、dbeableto:Slope:?1Intercept:?0Anytwopoints(AandB),oraninterceptandslope(?0and?1),definealineonatwo-dimensionalsurface.BAxyx2x1yCABAnythreepoints(A,B,andC),oraninterceptandcoefficientsofx1andx2(?0,?1?,and?2),defineaplaneinathree-dimensionalsurface.LinesPlanes11-1UsingStat

7、isticsThepopulationregressionmodelofadependentvariable,Y,onasetofkindependentvariables,X1,X2,...,Xkisgivenby:Y=?0+?1X1+?2X2+...+?kXk+?where?0istheY-interceptoftheregressionsurfaceandeach?i,i=1,2,...,kistheslopeoftheregressionsurface-sometimescalledtheresponsesurface-with

8、respecttoXi.x2x1y?2?1?0Modelassumptions:1.?~N(0,?2),independentofothererrors.2.ThevariablesXiareuncorre