A Few Useful Things to Know about Machine Learning

A Few Useful Things to Know about Machine Learning

ID:40556895

大小:139.89 KB

页数:9页

时间:2019-08-04

A Few Useful Things to Know about Machine Learning_第1页
A Few Useful Things to Know about Machine Learning_第2页
A Few Useful Things to Know about Machine Learning_第3页
A Few Useful Things to Know about Machine Learning_第4页
A Few Useful Things to Know about Machine Learning_第5页
资源描述:

《A Few Useful Things to Know about Machine Learning》由会员上传分享,免费在线阅读,更多相关内容在学术论文-天天文库

1、AFewUsefulThingstoKnowaboutMachineLearningPedroDomingosDepartmentofComputerScienceandEngineeringUniversityofWashingtonSeattle,WA98195-2350,U.S.A.pedrod@cs.washington.eduABSTRACTcorrectoutputytforfutureexamplesxt(e.g.,whethertheMachinelearningalgorithmscanfigureouthow

2、toperformspamfiltercorrectlyclassifiespreviouslyunseenemailsasimportanttasksbygeneralizingfromexamples.Thisisof-spamornotspam).tenfeasibleandcost-effectivewheremanualprogrammingisnot.Asmoredatabecomesavailable,moreambitious2.LEARNING=REPRESENTATION+problemscanbetackled

3、.Asaresult,machinelearningisEVALUATION+OPTIMIZATIONwidelyusedincomputerscienceandotherfields.However,Supposeyouhaveanapplicationthatyouthinkmachinedevelopingsuccessfulmachinelearningapplicationsrequireslearningmightbegoodfor.Thefirstproblemfacingyouasubstantialamounto

4、f“blackart”thatishardtofindinisthebewilderingvarietyoflearningalgorithmsavailable.textbooks.ThisarticlesummarizestwelvekeylessonsthatWhichonetouse?Thereareliterallythousandsavailable,machinelearningresearchersandpractitionershavelearned.andhundredsmorearepublishedeac

5、hyear.ThekeytonotTheseincludepitfallstoavoid,importantissuestofocuson,gettinglostinthishugespaceistorealizethatitconsistsandanswerstocommonquestions.ofcombinationsofjustthreecomponents.Thecomponentsare:1.INTRODUCTIONMachinelearningsystemsautomaticallylearnprogramsfr

6、omRepresentation.Aclassifiermustberepresentedinsomedata.Thisisoftenaveryattractivealternativetomanuallyformallanguagethatthecomputercanhandle.Con-constructingthem,andinthelastdecadetheuseofmachineversely,choosingarepresentationforalearneristan-learninghasspreadrapidl

7、ythroughoutcomputersciencetamounttochoosingthesetofclassifiersthatitcanandbeyond.MachinelearningisusedinWebsearch,spampossiblylearn.Thissetiscalledthehypothesisspacefilters,recommendersystems,adplacement,creditscoring,ofthelearner.Ifaclassifierisnotinthehypothesisfraud

8、detection,stocktrading,drugdesign,andmanyotherspace,itcannotbelearned.Arelatedquestion,whichapplications.ArecentreportfromtheMcKinseyGloba

当前文档最多预览五页,下载文档查看全文

此文档下载收益归作者所有

当前文档最多预览五页,下载文档查看全文
温馨提示:
1. 部分包含数学公式或PPT动画的文件,查看预览时可能会显示错乱或异常,文件下载后无此问题,请放心下载。
2. 本文档由用户上传,版权归属用户,天天文库负责整理代发布。如果您对本文档版权有争议请及时联系客服。
3. 下载前请仔细阅读文档内容,确认文档内容符合您的需求后进行下载,若出现内容与标题不符可向本站投诉处理。
4. 下载文档时可能由于网络波动等原因无法下载或下载错误,付费完成后未能成功下载的用户请联系客服处理。