●IntroductionbrPredictionVersusInterpretationbr2KeyIngredientsofPredictiveModelsbr3Terminologybr4ExampleDataSetsandTypicalDataScenariosbr5Overviewbr6NotationbrPartⅠGeneralStrategiesbr2AShortTourofthePredictiveModelingProcessbr2CaseStudyPredictingFuelEconomybr22Themesbr23Summarybr3DataPreprocessingbr3CaseStudyCellSegmentationinHighContentScreeningbr32DataTransformationsforIndividualPredictorsbr33DataTransformationsfor ltiplePredictorsbr34DealingwithMissingValuesbr35RemovingPredictorsbr36AddingPredictorsbr37BinningPredictorsbr38ComputingbrExercisesbr4OverFittingandModelTuningbr4TheProblemofOverFittingbr42ModelTuningbr43DataSplittingbr44ResamplingTechniquesbr45CaseStudyCreditScoringbr46ChoosingFinalTuningParametersbr47DataSplittingRecommendationsbr48ChoosingBetweenModelsbr49ComputingbrExercisesbrPartⅡRegressionModelsbr5MeasuringPerformanceinRegressionModelsbr5QuantitativeMeasuresofPerformancebr52TheVarianceBiasTradeoffbr53Computingbr6LinearRegressionandItsCousinsbr6CaseStudyQuantitativeStructureActivityRelationshirModelingbr62LinearRegressionbr63PartialLeastSquaresbr64PenalizedModelsbr65ComputingbrExercisesbr7NonlinearRegressionModelsbr7NeuralNetworksbr72 ltivariateAdaptiveRegressionSplinesbr73SupportVectorMachinesbr74KNearestNeighborsbr75ComputingbrExercisesbr8RegressionTreesandRuleBasedModelsbr8BasicRegressionTreesbr82RegressionModelTreesbr83RuleBasedModelsbr84BaggedTreesbr85RandomForestsbr86Boostingbr87Cubistbr88ComputingbrExercisesbr9ASummaryofSolubilityModelsbr0CaseStudyCompressiveStrengthofConcreteMixturesbr0ModelBuildingStrategybr02ModelPerformancebr03OptimizingCompressiveStrengthbr04ComputingbrPartⅢClassificationModelsbrMeasuringPerformanceinClassificationModelsbrClassPredictionsbr2EvaluatingPredictedClassesbr3EvaluatingClassProbabilitiesbr4Computingbr2DiscriminantAnalysisandOtherLinearClassificationModelsbr2CaseStudyPredictingSuccessfulGrantApplicationsbr22LogisticRegressionbr23LinearDiscriminantAnalysisbr24PartialLeastSquaresDiscriminantAnalysisbr25PenalizedModelsbr26NearestShrunkenCentroidsbr27ComputingbrExercisesbr3NonlinearClassificationModelsbr3NonlinearDiscriminantAnalysisbr32NeuralNetworksbr33FlexibleDiscriminantAnalysisbr34SupportVectorMachinesbr35KNearestNeighborsbr36NaiveBayesbr37ComputingbrExercisesbr4ClassificationTreesandRuleBasedModelsbr4BasicClassificationTreesbr42RuleBasedModelsbr43BaggedTreesbr44RandomForestsbr45Boostingbr46C50br47ComparingTwoEncodingsofCategoricalPredictorsbr48ComputingbrExercisesbr5ASummaryofGrantApplicationModelsbr6RemediesforSevereClassImbalancebr6CaseStudyPredictingCaravanPolicyOwnershipbr62TheEffectofClassImbalancebr63ModelTuningbr64AlternateCutoffsbr65AdjustingPriorProbabilitiesbr66UnequalCaseWeightsbr67SamplingMethodsbr68CostSensitiveTrainingbr69ComputingbrExercisesbr7CaseStudyJobSchedulingbr7DataSplittingandModelStrategybr72Resultsbr73ComputingbrPartⅣOtherConsiderationsbr8MeasuringPredictorImportancebr8NumericOutcomesbr82CategoricalOutcomesbr83OtherApproachesbr84ComputingbrExercisesbr9AnIntroductiontoFeatureSelectionbrbr

本書是一部關於數據分析的經典教材，聚焦預測建模的實際應用，如如何進行數據預處理、模型調優、預測變量重要性度量、變量選擇等。讀者可以從中學到許多建模方法以及提高對許多常用的、現代的有效模型的認識，如線性回歸、非線性回歸和分類模型，涉及樹方法、支持向量機等。書中還涉及從數據預處理到建模再到模型評估和選擇的整個過程，以及背後的統計思想，涉及各種回歸技術和分類技術。