Analysis:Googlemachinelearningscienceandtechnologybehindtheblack(FreepaperDownloadCenterNews)AfterusingtheAndroid(Andrews)open-sourcesystemchangethemobileworld,GooglehopestousetheopensourcemachinelearningsystemTensorFlowchangefieldofartificialintelligence.‘Machinelearningforme,istoenablehumanstomachinesthatdonotliketodothings.’November10,announcedthenextday,thesecond-generationopensourcemachinelearningsysteminGoogle,GoogleAlphabetparentcompanyChairmanEricSchmidt(EricSchmidt)indicatestoincludethe‘FirstFinancialDaily’throughthemedia,includingvideo.Facedwiththepresenceofthemedia,forexampleSchmidtsaid:‘Youareareporter,butalsokeptatthevenuewithacomputertyping,notjustlistenandsaywhynotletthemachinedothework,suchasduplicatemanufacturing,alotoftyping.work?‘Heexpectstoannounceasecond-generationopensourcemachinelearningsysteminGoogleafter,especiallyintheareasofhealthcare,gaming,educationan---本文于网络，仅供参考，勿照抄，如有侵权请联系删除---dsoon,machinelearningwillplayatremendousvalue.MachinelearningscienceandtechnologybehindtheblackMachineabilitytolearnjusthowimportant?GoogleresearchscientistGregCorradomadeametaphor:.‘Machinelearninglikerocketboosters,andalotofdataisrocketfuel.’Googlemachinelearningprincipleis:‘neuron’numerouscomputersimulationofthehumanbrain,theformationofaperson’s‘neuralnetwork’(ArtificialNeuralNetwork).Itdoesnotneedtheaidofalargenumberofresearcherstohelpcomputermarkeddifferencebetweenthings,aslongasthealgorithmprovidesvastamountsofdata,therelationshipbetween‘neurons’and‘neuron’willchangebetween,letthedataspeakforthemselves,sothatthecompositionof‘neuralnetworks‘machineswithautomaticlearning,theabilitytoidentifydatatoidentifyandconceptslearnedintheportioncorrespondingtothenewinput,tomachinelearning.Forexample,whenpeopleneedacomputertoidentifytheimagecontentwheneachartificialneuronwillpassthecapturedinformationtobesetto‘decision-makers’intheneurons,whichthroughitssystemconceptun---本文于网络，仅供参考，勿照抄，如有侵权请联系删除---derlyingallneuronspresentedinformation,combinedwithstories,dataanalysisandalgorithmfinalconclusion.Infact,Googleforresearchinmachinelearningdatesbacktothe7to8yearsago,voicetechnologyhasbegun.ButSchmidtsaid,machinelearningthistechnologybreakthrough,occurredinthefieldofcomputervision.Threeyearsago,GooglescientistJeffDean(JeffDean)inthe‘FirstFinancialDaily’interviewrevealed,‘GoogleX’laboratoriesbyconnecting16,000computerprocessors,createdamachinelearningneuralnetworksystem.Itwasfoundthatthesystemcreatesitsownconceptofthecatandthecat’sself-identification,whichis‘self-learning.’However,whentheGooglemachinelearningisonlyapilotproject,limitedtosimpletaskscognitiveclasses.Today,severalyearslater,Google’smachinelearninghasbeenupdatedfromtherecognitionGoogleapplicationlanguageandimagesofthefirstgenerationmachinelearningsystem‘DistBelief’tothesecondgenerationofTensorFlowsystem,andappliesGmail,GooglePhotos,Googletranslate,YouTubeandotherproducts.GoogleresearcherGregCorradotoldreporters,usingmachinelearningtechnology,thespame----本文于网络，仅供参考，勿照抄，如有侵权请联系删除---mailserviceGmailinterceptionrateincreasedto99.9%,thefalsepositiveratewasreducedto0.05%.Thereasonbehindthisisthatinthespamfilterintroducedmachinelearning,thistechnologycanlearntorecognizethrough...