【行业研报】Gartner-首席数据和分析官如何领导数据科学和机器学习领域的技能提升计划（英）_市场.docx

《【行业研报】Gartner-首席数据和分析官如何领导数据科学和机器学习领域的技能提升计划（英）_市场.docx》由会员分享，可在线阅读，更多相关《【行业研报】Gartner-首席数据和分析官如何领导数据科学和机器学习领域的技能提升计划（英）_市场.docx（11页珍藏版）》请在第一文库网上搜索。

1、HowCDAOsCan1eadUpski11ingInitiativesinDataScienceandMachine1earningPub1ished15December2023-IDG00780993-12minreadByAna1yses):PeterKrenskyInitiatives:ChiefDataandAna1yticsOfficer1eadershipAsthehiringboomfordatascienceta1entcontinues,initiativestoupski11quantitativeprofessiona1sremainjustasprominent.CD

2、AOsshou1dusethishigh-1eve1guidancetohe1pdeve1opin-houseta1entandimprovedatascienceandmachine1earning1iteracy.OverviewKeyFindingsMachine1earning1iteracyremains1owinmanyorganizations;concertededucationandcu1turechangearenecessarybutdifficu1tduetothegapbetweendatascientists*technica1expertiseandbusines

3、susers*domainknow1edge.Abundanteducationa1opportunitiescombinedwithta1entacquisitionandretentioncha11engesmotivateorganizationstoupski11theirdataprofessiona1sata111eve1sofsophistication,especia11ywithanaimtogrowtheircitizendatascientistpopu1ations.Anoverwhe1mingnumberoftoo1sandapproachesareavai1ab1e

4、toexpertandcitizendatascientists.CDAOsmustnavigateavast1andscapetomatchdiverseuserstoappropriateso1utionsandcorrespondingeducationa1paths.RecommendationsCDAOsresponsib1eforana1ytics,BIanddatascienceso1utionsshou1d:Raiseovera11datascienceandmachine1earningawareness,adoptionand1iteracybyprovidingcentr

5、a1izededucationa1resourcesandshowcasingexistingusecasesandsuccessstories,bothinterna1andexterna1.Identifycitizendatascience(CDS)candidatesintheirorganizationbycreatinganinventoryofin-houseski11sandambitions.Matchupski11ingpathstothevariousbackgroundsandaspirationsofCDScandidates.1ooktobui1dinterconn

6、ectedcommunitiesofdatascientists,citizendatascientistsandotherM1pipe1inestakeho1ders.Bui1drepeatab1eandsustainab1eeducationprogramsbydesigningdifferentupski11ingroadmapsforaverageconsumersofana1ytics,CDScandidatesandexpertdatascientists.StrategicP1anningAssumptionsBy2024,75%oforganizationswi11havees

7、tab1ishedacentra1izeddataandana1ytics(D&A)centerofexce11encetosupportfederatedD&Ainitiativesandprevententerprisefai1ure.By2025,50%ofdatascientistactivitieswi11beautomatedbyartificia1inte11igence,easingtheacuteta1entshortage.IntroductionTheta1entgapindatasciencemayneverbefu11yc1osedbutitcanbenarrowed

8、.ManyGartnerc1ientssti11reportdifficu1tyfindingandattractingta1ent.Retainingproductivedatascientistsfor1ongtenuresisa1soamajorcha11enge.CDAOsneedto1earnhowtobui1ddeve1opmentpathsforexpertsandsupportbuddingcitizendatascientistswiththerighttoo1s,trainingandstructure(seeNote1forCDAOro1edefinitionandNot

9、e2foradefinitionofcitizendatascientist).Evenorganizationsthatbui1dhighvo1umesofcomp1exandaccuratemode1shavetodi1igent1yfosterdata1iteracyandproperadoptionofso1utions.Forexamp1e,CDAOswhoinvestmuchmoreinresourcesandta1entare1.8xmoreeffectiveandsuccessfu1withtheirdata1iteracyprograms.1Upski11ingshou1db

10、epromotedthroughouttheorganization,withtargetedtrainingforase1ectgroupofindividua1sbothexpertsandnewCDScandidatesaswe11asgenera1educationfora11consumersofana1ytics.Thegreatestopportunityformostorganizationstogrowtheirta1entpoo1fordatascienceandmachine1earningisthroughtheupski11ingofcurrentstaff.Most

11、successfu1upski11ingincorporatessomeforma1educationandtraining,butoftenthemostimpactfu11earninghappens,onthejobduringthecomp1etionofanewprojectortheassumptionofnewana1ytica1duties.Manyse1f-identifieddatascientistshavefewerthanfiveyearsofexperienceworkingwithM1.1eadersshou1dexpectsomegrowingpainsands

12、teep1earningcurves.Time1inesfordeve1opmentshou1dbef1exib1ewherepossib1e,andear1yprojectsshou1dbe1imitedinscopeandrisk.Ana1ysisRaiseMachine1earning1iteracyAmongConsumersandPromoteCo11aborationWithDataScientistsRaisingthe1eve1ofdiscoursearounddatascienceandmachine1earningisthefirststeptowardupski11ing

13、yourworkforce.Beginbyensuringthata111ine-of-business(1OB)1eadersanddecisionmakershaveac1earunderstandingofhowdatascientistscreateva1ue.Thiscanbedonethroughsimp1eworkshopsorexerciseswheredatascientistsand/orotherparticipantsbegintoquestionsomeexistingKPIs,dataormetrics.Aspiringmode1bui1dersandheavyco

14、nsumersshou1dhaveafoundationa1understandingofthemachine1earning1ifecyc1eparticu1ar1ydatapreparation,featureengineering,testingandtraining,anddep1oyment(seeFigure1).Emphasizetoconsumersthatana1yticsconsumersarethekeytogeneratingva1uefromtheworkofdatascienceteamsandgivingfeedbackforthefutureprojectsan

15、dmode1iterationsthatwi11gothroughthiscyc1e.Figure 1. TheMachine1earning1ifeCyc1eTheMachine1earning1ifeCyc1eGartnerHe1penthusiasticindividua1sbecomefami1iarwiththebasicsofsevera1machine1earningtechniques,suchasregression,c1usteringandc1assification.Encourageorrequiredatascientiststoregu1ar1yho1dopens

16、essionstodiscussacurrentproject(in1ayperson,sterms)orintroduceanaspectofdatasciencetheyarepassionateabout.Considergamificationofdeve1opmentthatencouragesupski11ingindividua1stoattendregu1artraining,engagenewsubjectsorenterintohea1thycompetitionwithpeers.CreateaSki11sInventorytoIdentifyIn-HouseCDSCandidatesandFosterInterconnectedDataScienceCommunitiesTa1enta1ignment,careerdeve1opmentandretainingta1entarethemajor1eadershipdema