应用线性回归模型

出版时间：2005-2  出版社：蓝色畅想  作者：库特纳  页数：701  字数：1000000  
Tag标签：无  

内容概要

本书从McGrawHill出版公司引进，共分三部分，内容包括：第一部分：简单线性回归：一元预测函数的线性回归，回归影响和相关分析，诊断及补救措施，即时推断和回归分析的其它几个专题，简单线性回归分析中的矩阵方法；第二部分：多元线性回归：多元回归Ⅰ，多元回归2，定性回归模型和定量预测，建立线性回归模型Ⅰ：模型选择及有效性，建立线性回归模型Ⅱ：诊断，建立线性回归模型Ⅲ：补救措施，时间序列数据中的自相关；第三部分：非线性回归：非线性回归和神经网络方法。本书篇幅适中，例子多涉及各个应用领域，在介绍统计思想方面比较突出，光盘数据丰富。本书适用于高等院校统计学专业和理工科各专业本科生和研究生作为教材使用。

书籍目录

PARTONE　SIMPLELINEARREGRESSION.　Chapter1　LinearRegressionwithOnePredictorVariable　　1.1RelationsbetweenVariables　　1.2RegressionModelsandTheirUses　　1.3SimpleLinearRegressionModelwithDistributionofErrorTermsUnspecified　　1.4DataforRegressionAnalysis　　1.5OverviewofStepsinRegressionAnalysis　　1.6EstimationofRegressionFunction　　1.7EstimationofErrorTermsVarianceσ2　　1.8NormalErrorRegressionModel　Chapter2　InferencesinRegressionandCorrelationAnalysis　　2.1InferencesConcerning/β1　　2.2InferencesConcerning/β0　　2.3SomeConsiderationsonMakingInferencesConcerning/50andβ1　　2.4IntervalEstimationofE{Yh}　　2.5PredictionofNewObservation　　2.6ConfidenceBandforRegressionLine　　2.7AnalysisofVarianceApproach　　2.8GeneralLinearTestApproach　　2.9DescriptiveMeasuresofLinearAssociationbetweenXandY　　2.10ConsiderationsinApplyingRegressionAnalysis　　2.11NormalCorrelationModels　Chapter3　DiagnosticsandRemedialMeasures　　3.1DiagnosticsforPredictorVariable　　3.2Residuals　　3.3DiagnosticsforResiduals　　3.4OverviewofTestsInvolvingResiduals　　3.5CorrelationTestforNormality　　3.6TestsforConstancyofError　　3.7FTestforLackofFit　　3.8OverviewofRemedialMeasures　　3.9Transformations　　3.10ExplorationofShapeofRegressionFunction　　3.11CaseExample--PlutoniumMeasurement　Chapter4　SimultaneousInferencesandOtherTopicsinRegressionAnalysis　　4.1JointEstimationofβ0andβ1　　4.2SimultaneousEstimationofMeanResponses　　4.3SimultaneousPredictionIntervalsforNewObservations　　4.4RegressionthroughOrigin　　4.5EffectsofMeasurementErrors　　4.6InversePredictions　　4.7ChoiceofXLevels　Chapter5　MatrixApproachtoSimpleLinearRegressionAnalysis　　5.1Matrices　　5.2MatrixAdditionandSubtraction　　5.3MatrixMultiplication　　5.4SpecialTypesofMatrices　　5.5LinearDependenceandRankofMatrix　　5.6InverseofaMatrix　　5.7SomeBasicResultsforMatrices　　5.8RandomVectorsandMatrices　　5.9SimpleLinearRegressionModelinMatrixTerms　　5.10LeastSquaresEstimation　　5.11FittedValuesandResiduals　　5.12AnalysisofVarianceResults　　5.13InferencesinRegressionAnalysisPARTTWO　MULTIPLELINEARREGRESSION　Chapter6MultipleRegressionI　Chapter7　MultipleRegressionII　Chapter8　RegressionModelsforQuantitativeandQualitativePredictors　Chapter9　BuildingtheRegressionModelI:ModelSelectionandValidation　Chapter10　BuildingtheRegressionModelII:Diagnostics　Chapter11　BuildingtheRegressionModelIII:RemedialMeasures　Chapter12　AutocorrelationinTimeSeriesDataPARTTHREENONLINEARREGRESSION　Chapter13　IntroductiontoNonlinearRegressionandNeuralNetworks　Chapter14　LogisticRegression,PoissonRegression,andGeneralizedLinearModelsAppendixA　SomeBasicResultsinProbabilityandStatisticsAppendixB　TablesAppendixC　DataSetsAppendixD　SelectedBibliographyIndex

章节摘录

　　The correlation test for normality described in Chapter 3 carries forward directly to multipldregression.Tbe expected values of the ordered residuals under normality are calculatedaccording to（3.6），and the coefIicient of correlation between the residuals and the expectedvalues under normality is then obtained.Table B.6 is employed to assess whether or nolthe magnitude of the correlation coeIIicient supports the reasonableness of the normalityassumption.　　The Brown-Forsythe test statistic（3.9）for assessing the constancy ofthe error variance canbe used readily in multiple regression when the error variance increases or decreases withone of the predictor variables.To conduct the Brown-Forsythe teSt.we divide the data seinto two groups，as for simple linear regression，where one group consists of cases whenthe level of the predictor variable is relatively low and the other group consists of case where the level of the predictor variable is relatively hiRh.The Brown-Forsy the test the proceeds as for simple linear regression.The Breusch.Pagan test（3.1 1）for constancy of the error variance in multiple regression icarded out exactly the same as for simple linear regression when the error variance increaseor decreases with one of the predictor variables. 2. Research and Analysis （including site visit）　 A. Base Plan PreparationB. Site Inventory （Data Collection） and Analysis （Evaluation）C. Client InterviewD. Program Development

用户评论 (总计1条)

•   比我同学早两年买的同一本书的纸张要差一些。还没仔细看，希望没有其他问题吧。