ebook img

Practical Time Series Forecasting with R: A Hands-On Guide, 2nd Edition PDF

232 Pages·2016·4.31 MB·English
Save to my drive
Quick download
Download
Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.

Preview Practical Time Series Forecasting with R: A Hands-On Guide, 2nd Edition

G A L I T S H M U E L I K E N N E T H C. L I C H T E N DA H L J R . P R A C T I C A L T I M E S E R I E S F O R E C A S T I N G W I T H R A H A N D S - O N G U I D E SECOND EDITION AXELROD SCHNALL PUBLISHERS Copyright©2016GalitShmueli&KennethC.LichtendahlJr. published by axelrod schnall publishers isbn-13: 978-0-9978479-1-8 isbn-10: 0-9978479-1-3 Coverart: PunakhaDzong,Bhutan. Copyright©2016BoazShmueli ALLRIGHTSRESERVED.Nopartofthisworkmaybeusedorreproduced,transmitted, storedorusedinanyformorbyanymeansgraphic,electronic,ormechanical,includingbut notlimitedtophotocopying,recording,scanning,digitizing,taping,Webdistribution,infor- mationnetworksorinformationstorageandretrievalsystems,orinanymannerwhatsoever withoutpriorwrittenpermission. Forfurtherinformationseewww.forecastingbook.com SecondEdition,July2016 Contents Preface 9 1 ApproachingForecasting 15 1.1 Forecasting: Where? . . . . . . . . . . . . . . . . . . 15 1.2 BasicNotation . . . . . . . . . . . . . . . . . . . . . . 15 1.3 TheForecastingProcess . . . . . . . . . . . . . . . . 16 1.4 GoalDefinition . . . . . . . . . . . . . . . . . . . . . 18 1.5 Problems . . . . . . . . . . . . . . . . . . . . . . . . . 23 2 TimeSeriesData 25 2.1 DataCollection . . . . . . . . . . . . . . . . . . . . . 25 2.2 TimeSeriesComponents . . . . . . . . . . . . . . . . 28 2.3 VisualizingTimeSeries. . . . . . . . . . . . . . . . . 30 2.4 InteractiveVisualization . . . . . . . . . . . . . . . . 35 2.5 DataPre-Processing. . . . . . . . . . . . . . . . . . . 39 2.6 Problems . . . . . . . . . . . . . . . . . . . . . . . . . 42 3 PerformanceEvaluation 45 3.1 DataPartitioning . . . . . . . . . . . . . . . . . . . . 45 3.2 NaiveForecasts . . . . . . . . . . . . . . . . . . . . . 50 3.3 MeasuringPredictiveAccuracy . . . . . . . . . . . . 51 3.4 EvaluatingForecastUncertainty . . . . . . . . . . . 55 3.5 AdvancedDataPartitioning: Roll-ForwardValidation 62 3.6 Example: ComparingTwoModels . . . . . . . . . . 65 3.7 Problems . . . . . . . . . . . . . . . . . . . . . . . . . 67 4 ForecastingMethods: Overview 69 4.1 Model-Basedvs. Data-DrivenMethods . . . . . . . 69 4 4.2 ExtrapolationMethods,EconometricModels,andEx- ternalInformation . . . . . . . . . . . . . . . . . . . 70 4.3 Manualvs. AutomatedForecasting . . . . . . . . . 72 4.4 CombiningMethodsandEnsembles . . . . . . . . . 73 4.5 Problems . . . . . . . . . . . . . . . . . . . . . . . . . 77 5 SmoothingMethods 79 5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . 79 5.2 MovingAverage . . . . . . . . . . . . . . . . . . . . . 80 5.3 Differencing . . . . . . . . . . . . . . . . . . . . . . . 85 5.4 SimpleExponentialSmoothing . . . . . . . . . . . . 87 5.5 AdvancedExponentialSmoothing . . . . . . . . . . 90 5.6 SummaryofExponentialSmoothinginRUsingets 98 5.7 ExtensionsofExponentialSmoothing . . . . . . . . 101 5.8 Problems . . . . . . . . . . . . . . . . . . . . . . . . . 107 6 RegressionModels: Trend&Seasonality 117 6.1 ModelwithTrend . . . . . . . . . . . . . . . . . . . . 117 6.2 ModelwithSeasonality . . . . . . . . . . . . . . . . 125 6.3 ModelwithTrendandSeasonality . . . . . . . . . . 129 6.4 CreatingForecastsfromtheChosenModel . . . . . 132 6.5 Problems . . . . . . . . . . . . . . . . . . . . . . . . . 133 7 RegressionModels: Autocorrelation&ExternalInfo 143 7.1 Autocorrelation . . . . . . . . . . . . . . . . . . . . . 143 7.2 ImprovingForecastsbyCapturingAutocorrelation: ARandARIMAModels . . . . . . . . . . . . . . . . 147 7.3 EvaluatingPredictability . . . . . . . . . . . . . . . . 153 7.4 IncludingExternalInformation . . . . . . . . . . . . 154 7.5 Problems . . . . . . . . . . . . . . . . . . . . . . . . . 170 8 ForecastingBinaryOutcomes 179 8.1 ForecastingBinaryOutcomes . . . . . . . . . . . . . 179 8.2 NaiveForecastsandPerformanceEvaluation . . . . 180 8.3 LogisticRegression . . . . . . . . . . . . . . . . . . . 181 8.4 Example: RainfallinMelbourne,Australia . . . . . 183 8.5 Problems . . . . . . . . . . . . . . . . . . . . . . . . . 187 9 NeuralNetworks 189 5 9.1 NeuralNetworksforForecastingTimeSeries . . . . 189 9.2 TheNeuralNetworkModel . . . . . . . . . . . . . . 190 9.3 Pre-Processing . . . . . . . . . . . . . . . . . . . . . . 194 9.4 UserInput . . . . . . . . . . . . . . . . . . . . . . . . 195 9.5 ForecastingwithNeuralNetsinR . . . . . . . . . . 196 9.6 Example: ForecastingAmtrakRidership. . . . . . . 198 9.7 Problems . . . . . . . . . . . . . . . . . . . . . . . . . 201 10 CommunicationandMaintenance 203 10.1 PresentingForecasts . . . . . . . . . . . . . . . . . . 203 10.2 MonitoringForecasts . . . . . . . . . . . . . . . . . . 205 10.3 WrittenReports . . . . . . . . . . . . . . . . . . . . . 206 10.4 KeepingRecordsofForecasts . . . . . . . . . . . . . 207 10.5 AddressingManagerial"ForecastAdjustment" . . . 208 11 Cases 211 11.1 ForecastingPublicTransportationDemand . . . . . 211 11.2 ForecastingTourism(2010Competition,PartI) . . . 215 11.3 ForecastingStockPriceMovements(2010INFORMS Competition). . . . . . . . . . . . . . . . . . . . . . . 219 DataResources,Competitions,andCodingResources 225 Bibliography 227 Index 231 7 To Boaz Shmueli, who made the production of the Practical Analytics book series a reality Preface Thepurposeofthistextbookistointroducethereadertoquan- titativeforecastingoftimeseriesinapracticalandhands-on fashion. Mostpredictiveanalyticscoursesindatascienceand businessanalyticsprogramstouchverylightlyontimeseries forecasting,ifatall. Yet,forecastingisextremelypopularand usefulinpractice. Fromourexperience,learningisbestachievedbydoing. Hence,thebookisdesignedtoachieveself-learninginthefol- lowingways: • Thebookisrelativelyshortcomparedtoothertimeseries textbooks,toreducereadingtimeandincreasehands-ontime. • Explanationsstrivetobeclearandstraightforwardwithmore emphasisonconceptsthanonstatisticaltheory. • Chaptersincludeend-of-chapterproblems,ranginginfocus fromconceptualtohands-onexercises,withmanyrequiring runningsoftwareonrealdataandinterpretingtheoutputin lightofagivenproblem. • Realdataisusedtoillustratethemethodsthroughoutthe book. • Thebookemphasizestheentireforecastingprocessratherthan focusingonlyonparticularmodelsandalgorithms. • Casesaregiveninthelastchapter,guidingthereaderthrough suggestedsteps,butallowingself-solution. Workingonthe caseshelpsintegratetheinformationandexperiencegained. 10 Course Plan Thebookwasdesignedforaforecastingcourseatthegradu- ateorupper-undergraduatelevel. Itcanbetaughtinamini- semester(6-7weeks)orasasemester-longcourse,usingthe casestointegratethelearningfromdifferentchapters. Asug- gestedscheduleforatypicalcourseis: Week1 Chapters1("ApproachingForecasting")and2("Data") covergoaldefinition;datacollection,characterization,visualiza- tion,andpre-processing. Week2 Chapter3("PerformanceEvaluation")coversdatapar- titioning,naiveforecasts,measuringpredictiveaccuracyand uncertainty. Weeks3-4 Chapter4("ForecastingMethods: Overview")de- scribesandcomparesdifferentapproachesunderlyingforecast- ingmethods. Chapter5("SmoothingMethods")coversmoving average,exponentialsmoothing,anddifferencing. Weeks5-6 Chapters6("RegressionModels: TrendandSeason- ality")and7("RegressionModels: AutocorrelationandExternal Information")coverlinearregressionmodels,autoregressive (AR)andARIMAmodels,andmodelingexternalinformationas predictorsinaregressionmodel. Week7 Chapter10("CommunicationandMaintenance")dis- cussespracticalissuesofpresenting,reporting,documentingand monitoringforecasts. Thisweekisagoodpointforproviding feedbackonacaseanalysisfromChapter11. Week8(optional) Chapter8("ForecastingBinaryOutcomes") expandsforecastingtobinaryoutcomes,andintroducesthe methodoflogisticregression. Week9(optional) Chapter9("NeuralNetworks")introduces neuralnetworksforforecastingbothcontinuousandbinary outcomes.

Description:
PRACTICAL TIME SERIES FORECASTING WITH R: A HANDS-ON GUIDE, SECOND EDITION provides an applied approach to time-series forecasting. Forecasting is an essential component of predictive analytics. The book introduces popular forecasting methods and approaches used in a variety of business applications
See more

The list of books you might like

Most books are stored in the elastic cloud where traffic is expensive. For this reason, we have a limit on daily download.