
会员
Learn Amazon SageMaker
更新时间:2021-04-09 23:11:46 最新章节:Leave a review - let other readers know what you think
书籍简介
Quicklybuildanddeploymachinelearningmodelswithoutmanaginginfrastructure,andimproveproductivityusingAmazonSageMaker’scapabilitiessuchasAmazonSageMakerStudio,Autopilot,Experiments,Debugger,andModelMonitorKeyFeatures*Build,train,anddeploymachinelearningmodelsquicklyusingAmazonSageMaker*Analyze,detect,andreceivealertsrelatingtovariousbusinessproblemsusingmachinelearningalgorithmsandtechniques*Improveproductivitybytrainingandfine-tuningmachinelearningmodelsinproductionBookDescriptionAmazonSageMakerenablesyoutoquicklybuild,train,anddeploymachinelearning(ML)modelsatscale,withoutmanaginganyinfrastructure.IthelpsyoufocusontheMLproblemathandanddeployhigh-qualitymodelsbyremovingtheheavyliftingtypicallyinvolvedineachstepoftheMLprocess.ThisbookisacomprehensiveguidefordatascientistsandMLdeveloperswhowanttolearntheinsandoutsofAmazonSageMaker.You’llunderstandhowtousevariousmodulesofSageMakerasasingletoolsettosolvethechallengesfacedinML.Asyouprogress,you’llcoverfeaturessuchasAutoML,built-inalgorithmsandframeworks,andtheoptionforwritingyourowncodeandalgorithmstobuildMLmodels.Later,thebookwillshowyouhowtointegrateAmazonSageMakerwithpopulardeeplearninglibrariessuchasTensorFlowandPyTorchtoincreasethecapabilitiesofexistingmodels.You’llalsolearntogetthemodelstoproductionfasterwithminimumeffortandatalowercost.Finally,you’llexplorehowtouseAmazonSageMakerDebuggertoanalyze,detect,andhighlightproblemstounderstandthecurrentmodelstateandimprovemodelaccuracy.BytheendofthisAmazonbook,you’llbeabletouseAmazonSageMakeronthefullspectrumofMLworkflows,fromexperimentation,training,andmonitoringtoscaling,deployment,andautomation.Whatyouwilllearn*Createandautomateend-to-endmachinelearningworkflowsonAmazonWebServices(AWS)*Becomewell-versedwithdataannotationandpreparationtechniques*UseAutoMLfeaturestobuildandtrainmachinelearningmodelswithAutoPilot*Createmodelsusingbuilt-inalgorithmsandframeworksandyourowncode*TraincomputervisionandNLPmodelsusingreal-worldexamples*Covertrainingtechniquesforscaling,modeloptimization,modeldebugging,andcostoptimization*AutomatedeploymenttasksinavarietyofconfigurationsusingSDKandseveralautomationtoolsWhothisbookisforThisbookisforsoftwareengineers,machinelearningdevelopers,datascientists,andAWSuserswhoarenewtousingAmazonSageMakerandwanttobuildhigh-qualitymachinelearningmodelswithoutworryingaboutinfrastructure.KnowledgeofAWSbasicsisrequiredtograsptheconceptscoveredinthisbookmoreeffectively.SomeunderstandingofmachinelearningconceptsandthePythonprogramminglanguagewillalsobebeneficial.
品牌:中图公司
上架时间:2020-08-27 00:00:00
出版社:Packt Publishing
本书数字版权由中图公司提供,并由其授权上海阅文信息技术有限公司制作发行
最新章节
Julien Simon;Francesco Pochetti
同类热门书
最新上架
同类书籍最近更新
- 会员Thisbookisastep-by-steptutorialfilledwithpracticalexampleswhichwillfocusmainlyonthekeysecuritytoolsandimplementationtechniquesofHadoopsecurity.ThisbookisgreatforHadooppra统计2.5万字
- 会员本书将着重对项目族工作分解结构模型与配置和多项目进度优化模型与算法展开研究与探索。本专著共分为六个部分,计15章。第一部分主要介绍多项目管理研究概况,主要目的是为后续研究作背景铺垫。第二部分着重阐述项目族工作分解结构模型构建与配置机理。该部分对项目族内WBS元素进行归并、标准化及WBS词典属性化,建立了项目族工作分解结构。第三部分从项目族工作分解结构应用角度,分别介绍项目族工作分解结构在模型构建、统计20.8万字
- 会员ThisbookisgearedupforanalystsandaccountantskeenonbuildingandmaintainingprofessionaldashboardswithMicrosoftExcel2016forMicrosoftDynamicsGP2016data,andbuildingfinancialswith统计6.2万字