會(huì)計(jì)數(shù)據(jù)分析 課件 英文 第3、4章　模型建立與評(píng)估、可視化：使用可視化工具和書面報(bào)告等向利益相關(guān)者分享分析結(jié)果

ModelingandEvaluationChapter3Wherewearenow1.DataAnalytics2.DataPreparationandCleaning3.ModelingandEvaluation4.Visualization5.TheModernAudit6.AuditAnalytics7.KeyPerformanceIndicators8.FinancialStatementAnalyticsObjectivesLO3-1Whatdoyouneedtoknowaboutdatamodels?LO3-2Whataresomeunsupervisedandsupervisedapproaches?LO3-3Howdoyouperformprofiling?LO3-4Howdoyouperformdatareduction?LO3-5Howdoyouperformregression?LO3-6Howdoyouperformclassification?LO3-7Howdoyouperformclustering?IntheIMPACTcycle,we’renowgoingtolookatPerformingtheTestPlanandAddressingtheResults.IdentifythequestionsMasterthedataPerformtestplanAddressandrefineresultsCommunicateinsightsTrackoutcomesExhibit1-1TheIMPACTCycleWhatdoyouneedtoknowaboutdatamodels?LO3-1Whatdoyouneedtoknowaboutdatamodels?Atargetisanexpectedattributeorvaluethatwewanttoevaluate.Example:FraudscoreInterestrateAclassisamanuallyassignedcategoryappliedtoarecordbasedonanevent.Example:Accept/RejectFraud/NotfraudWhatdoyouneedtoknowfirst?Anunsupervisedapproachisusedwhenyoudon’thaveaspecificquestion.Example:“Doourvendorsformnaturalgroupsbasedonsimilarattributes?”Asupervisedapproach

isusedwhenyouaretryingtopredictafutureoutcomebasedonhistoricaldata.Example:“Willanewvendorshipalargeorderontime?”Whataresomeunsupervisedapproaches?Clustering

–findundiscoverednaturalgroupingsinthedataCo-occurrencegrouping–eventsthathappentogetherProfiling

–identifytypicalbehaviorinthedataDatareduction–filterorgroupthedatatosimplifytheanalysisWhataresomesupervisedapproaches?Classification

–predictwhetherdatabelongstooneclassoranotherSimilaritymatching–groupdatabyattributesRegression

–predictaspecificvalueLinkprediction–socialnetworksCausalmodeling–aneventinfluencesanotherUseaflowcharttoidentifyanappropriateapproach.Q.Whatisthemaindifferencebetweensupervisedandunsupervisedmethods?Howdoyouperformprofiling?LO3-2Profilingreliesongatheringsummarystatisticsandidentifyingoutliers.Identifytheobjectsoractivityyouwanttoprofile.Determinethetypesofprofilingyouwanttoperform.Setboundariesorthresholdsfortheactivity.Interprettheresultsandmonitortheactivityand/orgeneratealistofexceptions.Followuponexceptions.Whataresomeexamplesofprofiling?Internalauditorsanalyzetravelandentertainmentexpensesforviolationsofinternalcontrols.Managersuseprofilingtocomparevariancesfromtargetranges.Whataresomeexamplesofprofiling?Inthecontinuousaudit,anauditormayuseBenford’sLawtoevaluatethefrequencydistributionofthefirstdigitsfromalargesetofnumericaldata.Q.Profilingisusedinlawenforcementforoffenderorcriminalprofiling.Howdoesthiscomparewithprofilingaccountingdata?Howdoyouperformdatareduction?LO3-3Datareductionisusedtofilterresults.1.Identifytheattributeyouwouldliketoreduceorfocuson.2.Filtertheresults.3.Interprettheresults.4.Followupontheresults.Whataresomeexamplesofdatareduction?InternalauditorsmaywanttolocatepaymentsmadetoSquarevendors.FinancialstatementanalystswilltakeXBRLinstancedocumentsandfilteronspecifictags.Q.HowmightthedatareductionapproachbeusedtosimplifyT&Eexpenses?Howdoyouperformregression?LO3-4Regressionallowstheaccountanttodevelopmodelstopredictexpectedoutcomes.Identifythevariablesthatmightpredictanoutcome.Determinethefunctionalformoftherelationship.Identifytheparametersofthemodel.Dependentvariable=f(independentvariables)Whataresomeexamplesofregression?Inmanagerialaccounting,regressionmaypredictemployeeturnover:Employeeturnover=f(currentprofessionalsalaries,healthoftheeconomy[GDP],salariesofferedbyotheraccountingfirmsorbycorporateaccounting,etc.)Inauditing,regressionmaybeusedtodeterminetheappropriatenessofallowanceaccounts:Allowanceforloanlosesamount=f(currentagedloans,loantype,customerloanhistory,collectionssuccess)Q.Regressionisusedtopredictstockreturnsfollowingarestatementofpastearnings.Whatfactors(independentvariables)doyouthinkmightpredictthechangeinstockprice(dependentvariable)?Howdoyouperformclassification?LO3-5Thegoalofclassificationistopredictwhetheranindividualweknowverylittleaboutwillbelongtooneclassoranother.Identifytheclassesyouwishtopredict.Manuallyclassifyanexistingsetofrecords.Selectasetofclassificationmodels.Divideyourdataintotrainingandtestingsets.Generateyourmodel.Interprettheresultsandselectthe“best”model.Whatelsedoyouneedtoknowaboutclassification?Trainingdataareexistingdatathathavebeenmanuallyevaluatedandassignedaclass.Testdataareexistingdatausedtoevaluatethemodel.Decisiontreesareusedtodividedataintosmallergroups.Decisionboundariesmarkthesplitbetweenoneclassandanother.Whatelsedoyouneedtoknowaboutclassification?Pruning

removesbranchesfromadecisiontreetoavoidoverfittingthemodel.Whatelsedoyouneedtoknowaboutclassification?Linearclassifiersareusefulforrankingitemsratherthansimplypredictingclassprobability.Theseareusefulfordeterminingthereallyimportantvalues,suchasvaluablecustomers,orwhichtransactionsaremostlikelyfraudulent.Whatelsedoyouneedtoknowaboutclassification?Supportvectormachineisadiscriminatingclassifierthatisdefinedbyaseparatinghyperplanethatworksfirsttofindthewidestmargin(orbiggestpipe)andthenworkstofindthemiddleline.Howdoweevaluateclassifiers?Trytoavoidoverfitting,ormodelsthataretooaccurate.Theyareactuallyprettybadapredictingafutureobservation.Lookforthesweetspotwherewemaximizetheaccuracyofthetestingdata.Q.Ifwearetryingtopredictwhetheraloanwillberejected,wouldyouexpectcreditscoretobepositivelyornegativelyassociatedwithloanrejection?Howdoyouperformclustering?LO3-6Howdoyouperformclustering?Clusteringisusedtoidentifygroupsofsimilardataelementsandtheunderlyingdriversofthosegroups.Clusteringalgorithmscalculatetheminimumdistanceofallobservationsandgroupsthoseelements.Whataresomeexamplesofclustering?Internalauditorscanuseclusteringtoidentifygroupsoftransactionsthatmayindicateriskorfraudininsuranceorotherpayments.Q.WhatarethreeclustersofcustomerswhomightshopatWalmart?SummaryInthischapter,weaddressedthethirdstepoftheIMPACTcyclemodel:the“P”for“performingtestplan.”Thatis,howarewegoingtotestoranalyzethedatatoaddressaproblemwearefacing?Basedonourproblemandthedataavailable,weprovidedaflowchartthathelpstheanalysttochoosethemostappropriatemodel,notingthedifferenceswhenweuseasupervisedversusanunsupervisedapproach.Specifically,weaddressedfivedataanalyticsapproachesortechniquesaremostcommontoaddressouraccountingquestions:profiling,datareduction,regression,classification,andclustering.Wealsoprovidedexamplesofaccountingandauditingproblemsaddressedbythesedataapproaches.WeintroducedtheconceptsofBenford’slawandfuzzymatch,whichwewilluseinsubsequentchapters.Wepresentedsomeclassificationterminology—includingtestandtrainingdata,decisiontreesandboundaries,linearclassifiers,andsupportvectormachines—andtalkedabouttheperilsofunder-andoverfittingthetrainingdataanditsconsequencesinpredictionsusingthetestdata.Visualization:UsingVisualizationsandSummariestoShareResultswithStakeholdersChapter4Wherewearenow1.DataAnalytics2.DataPreparationandCleaning3.ModelingandEvaluation4.Visualization5.TheModernAudit6.AuditAnalytics7.KeyPerformanceIndicators8.FinancialStatementAnalyticsObjectivesLO4-1DeterminethepurposeofyourdatavisualizationLO4-2ChoosethebestchartforyourdatasetLO4-3RefineyourcharttocommunicateefficientlyandeffectivelyLO4-4CommunicateyourresultsinawrittenreportIntheIMPACTcycle,we’renowgoingtolookatCommunicatingInsightsandTrackingOutcomes.IdentifythequestionsMasterthedataPerformtestplanAddressandrefineresultsCommunicateinsightsTrackoutcomesExhibit1-1TheIMPACTCycleDataAnalyticsareeffective,buttheyareonlyasimportantandeffectiveaswecancommunicateandmakethedataunderstandable.Whatisthepurposeofyourdatavisualization?LO4-1Whattypeofdataisbeingvisualized?

Areyouexplainingresultsorexploringthedata?Exhibit4-2ThefourcharttypesAreyouusingqualitativeandquantitativedata?Qualitativedataarecategoricaldata

(e.g.count,group,rank)Nominaldataissimple.

(e.g.haircolor)Ordinaldatacanberanked.

(e.g.gold,silver,bronze)Proportionshowsthemakeupofeachcategory.

(e.g.55%cats,45%dogs)Quantitativedataarenumerical

(e.g.age,height,dollaramount)Ratiodatadefines0as“absenceof”something.(e.g.cash)Intervaldatawhere0isjustanothernumber.(e.g.temperature)Discretedatashowonlywholenumbers.(e.g.pointsinabasketballgame)Continuousdatashownumberswithdecimals.(e.g.height)Distributions

describethemean,median,andstandarddeviationofthedata.Isyourvisualizationdeclarativeorexploratory?Declarativevisualizationsareusedtopresentfindings.

(e.g.financialresults)

Exploratoryvisualizationsareusedtogaininsightswhileyouareinteractingwithdata.

(e.g.identifyinggoodcustomers)Onceyouhavedefinedyourdataandthepurpose,youcanfindanappropriatechartorgraph.Exhibit4-3ThefourcharttypeswithdetailsQ.Whichtypeofdatascaleshouldthefollowingvariablesbemeasuredon?Instructorevaluations(excellent,good,average,poor)WeeklyclosingpriceofgoldNamesofcompanieslistedontheDowJonesIndustrialAverageFahrenheitscaleformeasuringtemperatureHowdoyouchoosetherightchart?LO4-2Whichchartsareappropriateforqualitativedata?Whenyouwanttoshowproportion:BarchartsPiechartsStackedbarchartTreemapsHeatmapsSymbolmapsWordcloudsWhichchartsareappropriateforquantitativedata?Whenyouwanttoshowcomplexdata:LinechartsBoxandwhiskerplotsScatterplotsFilledgeographicmapsHereisasummaryguideofwhentousedifferentvisualizations.

AlsocheckoutConceptual(Qualitative)Comparison:BarchartPiechartStackedbarchartTreemapHeatmapGeographicdata:SymbolmapTextdata:WordcloudData-Driven(Quantitative)Outlierdetection:BoxandwhiskerplotRelationshipbetweentwovariables:ScatterplotTrendovertime:

LinechartGeographicdata:

FilledmapExhibit4-8TypesofchartsWhichtoolsarehelpfulforcreatingvisualizations?TableauandMicrosoftBIaregreatforexploratorydataanalysis.TableauandMicrosoftBItopthelistofvisionaryleadersforvisualizationtools.MicrosoftExcelisgoodforbasicdeclarativecharts.Exhibit4-9GartnerMagicQuadrantforBusinessIntelligenceandAnalyticsPlatformsBadexample:Howdoesthischartillustratebias?Howbigofachangedoesthisrepresent?Whymightthecreatormakethischart?Exhibit4-12Amoreappropriatescaleisagoodstart.Exhibit4-12Exhibit4-13Stackingcanrevealtherealincrease.Exhibit4-12Exhibit4-14Badexample:Whatisthischarttryingtotellthereaderaboutwhosecomputerisattackedmore?IsFinnaScientist?IsMarkusanAdministrator?Dowecaremoreaboutthepeopleorroles?Howdotheusersevencompare?Exhibit4-15Ifwecareaboutindividuals,anorderedbarchartisalittlemoreclear.Exhibit4-15Exhibit4-16Ifwecareaboutfunction,anbarchartcanshowtheproportionmoreclearly.Exhibit4-15Exhibit4-17Andastackedbarchartisalmostalwayseasiertointerpret(inlessspace)thanapie.Exhibit4-15Exhibit4-18Q.Whichcharttypeisbetterforworkingwithdatesovertime,abarchartorlinegraph?Why?Howcanyourefineyourcharts?LO4-3Improvingyourchartscomesdowntochoosinganappropriatescaleandusingcolorseffectively.Considerscaleandincrements:Howmuchdatadoyouneedtoshow?Whatdoyoudowithoutliers?Whatisthebaseline?0?Somethingelse?Wouldcontextorreferencelinesmakethescalemoremeaningful?Thinkaboutyouruseofcolor:Whatdothecolorsmean?Shouldredbeusedforpositiveoutcomes?Whatcolorschemewouldhelpyourcolor-blindparticipants?Q.Howmighttheuseofcompanycolorsorlogosdistractfromavisualization?Howcantheuseofwordsprovideinsight?LO4-4Gettothepoint.Beclear,unambiguous,correct,interesting,anddirect.RemembertouseplainlanguagethroughouttheIMPACTmodelI:Explainwhatwasbeingresearchedandthepurposeoftheproject.M:Ifappropriate,describeissuesyouencounteredintheETLprocess.PandA:Giveanoverviewofyourmodelandlimitationsyoufaced.C:Provideanexplanationofthevisualyouchose.Describeanyitemsthatstandoutorthatareinteresting.T:Discusswhat’snextinyouranalysis.Howfrequentlywillitbeupdated?Aretheretrendsoroutliersthatshouldbepaidattention?ConsideryouraudienceandtonePlacethefocusonyouraudience.Craftdifferentversionsfordifferentaudiences.Useanappropriate

tone.Providetherightcontent.Avoidtoomuchdetail.Don’tforgettoreviseasneeded.Askotherpeopletoreadthoughyourwritingtomakesureyouareclear.

Q:IfyouwerepresentingdataonsalesforacompanylikeSláinte,howwouldyoudescribetheETLprocesstotheCEO?Totheprogrammerscreatingthereport?SummaryThischapterfocusedonthefifthstepoftheIMPACTmodel,orthe“C,”todiscusshowtocommunicatetheresultsofyourdataanalysisprojects.Communicationcanbedonethroughavarietyofdatavisualizationsandwrittenreports,dependingonyouraudienceandthedatayouareexhibiting.nordertoselecttheri