PrefaceIwrotethisbooktohelpmachinelearningpractitioners,likeyou,getontopoflinearalgebra,fast.LinearAlgebraIsImportantinMachineLearningThereisnodoubtthatlinearalgebraisimportantinmachinelearning.Linearalgebraisthemathematicsofdata.It’sallvectorsandmatricesofnumbers.Modernstatisticsisdescribedusingthenotationoflinearalgebraandmodernstatisticalmethodsharnessthetoolsoflinearalgebra.Modernmachinelearningmethodsaredescribedthesameway,usingthenotationsandtoolsdrawndirectlyfromlinearalgebra.Evensomeclassicalmethodsusedinthefield,suchaslinearregressionvialinearleastsquaresandsingular-valuedecomposition,arelinearalgebramethods,andothermethods,suchasprincipalcomponentanalysis,werebornfromthemarriageoflinearalgebraandstatistics.Toreadandunderstandmachinelearning,youmustbeabletoreadandunderstandlinearalgebra.PractitionersStudyLinearAlgebraTooEarlyIfyouaskhowtogetstartedinmachinelearning,youwillverylikelybetoldtostartwithlinearalgebra.Weknowthatknowledgeoflinearalgebraiscriticallyimportant,butitdoesnothavetobetheplacetostart.Learninglinearalgebrafirst,thencalculus,probability,statistics,andeventuallymachinelearningtheoryisalongandslowbottom-uppath.Abetterfitfordevelopersistostartwithsystematicproceduresthatgetresults,andworkbacktothedeeperunderstandingoftheory,usingworkingresultsasacontext.Icallthisthetop-downorresults-firstapproachtomachinelearning,andlinearalgebraisnotthefirststep,butperhapsthesecondorthird.PractitionersStudyTooMuchLinearAlgebraWhenpractitionersdocirclebacktostudylinearalgebra,theylearnfarmoreofthefieldthanisrequiredfororrelevanttomachinelearning.Linearalgebraisalargefieldofstudythathastendrilsintoengineering,physicsandquantumphysics.Therearealso

InATourofC++,SecondEdition,BjarneStroustrup,thecreatorofC++,describeswhatconstitutesmodernC++.Thisconcise,self-containedguidecoversmostmajorlanguagefeaturesandthemajorstandard-librarycomponents—not,ofcourse,ingreatdepth,buttoalevelthatgivesprogrammersameaningfuloverviewofthelanguage,somekeyexamples,andpracticalhelpingettingstarted.StroustruppresentstheC++featuresinthecontextoftheprogrammingstylestheysupport,suchasobject-orientedandgenericprogramming.Histourisremarkablycomprehensive.Coveragebeginswiththebasics,thenrangeswidelythroughmoreadvancedtopics,includingmanythatarenewinC++17,suchasmovesemantics,uniforminitialization,lambdaexpressions,improvedcontainers,randomnumbers,andconcurrency.ThetourevencoverssomeextensionsbeingmadeforC++20,suchasconceptsandmodules,andendswithadiscussionofthedesignandevolutionofC++.Thisguidedoesnotaimtoteachyouhowtoprogram(forthat,seeStroustrup’sProgramming:PrinciplesandPracticeUsingC++,SecondEdition),norwillitbetheonlyresourceyou’llneedforC++mastery(forthat,seeStroustrup’sTheC++ProgrammingLanguage,FourthEdition,andrecommendedonlinesources).If,however,youareaCorC++programmerwantinggreaterfamiliaritywiththecurrentC++language,oraprogrammerversedinanotherlanguagewishingtogainanaccuratepictureofthenatureandbenefitsofmodernC++,youcan’tfindashorterorsimplerintroductionthanthistourprovides.

WelcometoLongShort-TermMemoryNetworksWithPython.LongShort-TermMemory(LSTM)recurrentneuralnetworksareoneofthemostinterestingtypesofdeeplearningatthemoment.Theyhavebeenusedtodemonstrateworld-classresultsincomplexproblemdomainssuchaslanguagetranslation,automaticimagecaptioning,andtextgeneration.LSTMsareverydi↵erenttootherdeeplearningtechniques,suchasMultilayerPerceptrons(MLPs)andConvolutionalNeuralNetworks(CNNs),inthattheyaredesignedspecificallyforsequencepredictionproblems.IdesignedthisbookforyoutorapidlydiscoverwhatLSTMsare,howtheywork,andhowyoucanbringthisimportanttechnologytoyourownsequencepredictionproblems.

WelcomeWelcometoXGBoostWithPython.ThisbookisyourguidetofastgradientboostinginPython.YouwilldiscovertheXGBoostPythonlibraryforgradientboostingandhowtouseittodevelopandevaluategradientboostingmodels.Inthisbookyouwilldiscoverthetechniques,recipesandskillswithXGBoostthatyoucanthenbringtoyourownmachinelearningprojects.GradientBoostingdoeshaveasomefascinatingmathunderthecovers,butyoudonotneedtoknowittobeabletopickitupasatoolandwielditonimportantprojectstodeliverrealvalue.Fromtheappliedperspective,gradientboostingisquiteashallowfieldandamotivateddevelopercanquicklypickitupandstartmakingveryrealandimpactfulcontributions.Thisismygoalforyouandthisbookisyourtickettothatoutcome.

在日常工作中，钉钉打卡成了我生活中不可或缺的一部分。然而，有时候这个看似简单的任务却给我带来了不少烦恼。 每天早晚，我总是得牢记打开钉钉应用，点击"工作台"，再找到"考勤打卡"进行签到。有时候因为工作忙碌，会忘记打卡，导致考勤异常，影响当月的工作评价。而且，由于我使用的是苹果手机，有时候系统更新后，钉钉的某些功能会出现异常，使得打卡变得更加麻烦。 另外，我的家人使用的是安卓手机，他们也经常抱怨钉钉打卡的繁琐。尤其是对于那些不太熟悉手机操作的长辈来说，每次打卡都是一次挑战。他们总是担心自己会操作失误，导致打卡失败。 为了解决这些烦恼，我开始思考是否可以通过编写一个全自动化脚本来实现钉钉打卡。经过一段时间的摸索和学习，我终于成功编写出了一个适用于苹果和安卓系统的钉钉打卡脚本。