人工智能英語(yǔ)教程參考試卷

參考試卷

一、寫(xiě)出以下單詞的中文意思（每小題0.5分，共10分）

1accuracy11customize

2actuator12definition

3adjust13defuzzification

4agent14deployment

5algorithm15effector

6analogy16entity

7attribute17extract

8backtrack18feedback

9blockchain19finite

10cluster20framework

二、根據(jù)給出的中文意思，寫(xiě)出英文單詞（每小題0.5分，共10分）

1V.收集，搜集11n.神經(jīng)元；神經(jīng)細(xì)胞

2adj.嵌入的，內(nèi)置的12n.節(jié)點(diǎn)

3n.指示器；指標(biāo)13V.運(yùn)轉(zhuǎn)；操作

4n.基礎(chǔ)設(shè)施，基礎(chǔ)架構(gòu)14n.模式

5V.合并：集成15V.察覺(jué)，發(fā)覺(jué)

6n.解釋器，解釋程序16n.前提

7n.迭代；循環(huán)17adj.程序的；過(guò)程的

8n.庫(kù)18n.回歸

919adj.健壯的，強(qiáng)健的;

n.元數(shù)據(jù)

結(jié)實(shí)的

10v.監(jiān)視；控制；監(jiān)測(cè)20V.篩選

三、根據(jù)給出的短語(yǔ)，寫(xiě)出中文意思（每小題1分，共10分）

1dataobject

2cybersecurity

3smartmanufacturing

4clusteredsystem

5datavisualization

6opensource

7analyzetext

8cloudcomputing

9computationpower

10objectrecognition

四、根據(jù)給出的中文意思，寫(xiě)出英文短語(yǔ)（每小題1分，共10分）

1數(shù)據(jù)結(jié)構(gòu)______________________

2決策樹(shù)______________________

3演繹推理______________________

4貪婪最佳優(yōu)先搜索______________________

5隱藏模式，隱含模式______________________

6知識(shí)挖掘______________________

7邏輯推理______________________

8預(yù)測(cè)性維護(hù)______________________

9搜索引擎______________________

10文本挖掘技術(shù)

五、寫(xiě)出以下縮略語(yǔ)的完整形式和中文意思（每小題1分，共10分）

縮略語(yǔ)_______________完整形式中文意思___________

1ANN

2AR

3BFS

4CV

5DFS

6ES

7IA

8KNN

9NLP

10VR

六、閱讀短文，回答問(wèn)題(每小題2分，共10分)

ArtificialNeuralNetwork(ANN)

Anartificialneuralnetwork(ANN)isthepieceofacomputingsystemdesignedtosimulate

thewaythehumanbrainanalyzesandprocessesinformation.Itisthefoundationofartificial

intelligence(AI)andsolvesproblemsthatwouldproveimpossibleordifficultbyhumanor

statisticalstandards.ANNshaveself-learningcapabilitiesthatenablethemtoproducebetter

resultsasmoredatabecomesavailable.

Artificialneuralnetworksarebuiltlikethehumanbrain,withneuronnodesinterconnected

likeaweb.Thehumanbrainhashundredsofbillionsofcellscalledneurons.Eachneuronismade

upofacellbodythatisresponsibleforprocessinginformationbycarryinginformationtowards

(inputs)andaway(outputs)fromthebrain.

AnANNhashundredsorthousandsofartificialneuronscalledprocessingunits,whichare

interconnectedbynodes.Theseprocessingunitsaremadeupofinputandoutputunits.Theinput

unitsreceivevariousformsandstructuresofinformationbasedonaninternalweightingsystem,

andtheneuralnetworkattemptstolearnabouttheinformationpresentedtoproduceoneoutput

report.Justlikehumansneedrulesandguidelinestocomeupwitharesultoroutput,ANNsalso

useasetoflearningrulescalledbackpropagation,anabbreviationfbrbackwardpropagationof

error,toperfecttheiroutputresults.

AnANNinitiallygoesthroughatrainingphasewhereitlearnstorecognizepatternsindata,

whethervisually,aurally,ortextually.Duringthissupervisedphase,thenetworkcomparesits

actualoutputproducedwithwhatitwasmeanttoproduce—thedesiredoutput.Thedifference

betweenbothoutcomesisadjustedusingbackpropagation.Thismeansthatthenetworkworks

backward,goingfromtheoutputunittotheinputunitstoadjusttheweightofitsconnections

betweentheunitsuntil(hedifferencebetweentheactualanddesiredoutcomeproducesthelowest

possibleerror.

Aneuralnetworkmaycontainthefollowing3layers:

Inputlayer-Theactivityoftheinputunitsrepresentstherawinformationthatcanfeedinto

thenetwork.

Hiddenlayer-Todeterminetheactivityofeachhiddenunit.Theactivitiesoftheinputunits

andtheweightsontheconnectionsbetweentheinputandthehiddenunits.Theremaybeoneor

morehiddenlayers.

Outputlayer-Thebehavioroftheoutputunitsdependsontheactivityofthehiddenunits

andtheweightsbetweenthehiddenandoutputunits.

1.Whatisanartificialneuralnetwork(ANN)?

2.Whatiseachneuronmadeupof?

3.Whadotheinputunitsdo?

4.WhatdoesanANNinitiallygothrough?

5.Howmanylayersmayaneuralnetworkcontain?Whatarethey?

七、將下列詞填入適當(dāng)?shù)奈恢?每詞只用一次)。(每小題10分，共20分)

填空題1

供選擇的答案:

transactionsinformationtechniquesfraudnodes

unstructuredsubsetsharedautomatedexplosion

DeepLearning

1.WhatIsDeepLearning?

Deeplearningisanartificialintelligence(AI)functionthatimitatestheworkingsofthe

humanbraininprocessingdataandcreatingpatternsforuseindecisionmaking.Deeplearningis

a___1___ofmachinelearninginartificialintelligencethathasnetworkscapableoflearning

unsupervisedfromdatathatis___2___orunlabeled.Alsoknownasdeepneurallearningordeep

neuralnetwork.

2.HowDoesDeepLearningWork?

Deeplearninghasevolvedhand-in-handwiththedigitalera,whichhasbroughtaboutan

___3___ofdatainallformsandfromeveryregionoftheworld.Thisdata,knownsimplyasbig

data,isdrawnfromsourceslikesocialmedia,internetsearchengines,e-commerceplatforms,and

onlinecinemas,amongothers.Thisenormousamountofdataisreadilyaccessibleandcanbe

___4___throughfintechapplicationslikecloudcomputing.

However,thedata,whichnormallyisunstructured,issovastthatitcouldtakedecadesfor

humanstocomprehenditandextractrelevant___5___.Companiesrealizetheincrediblepotential

thatcanresultfromunravelingthiswealthofinformationandareincreasinglyadaptingtoAI

systemsfor___6___support.

3.DeepLearningvs.MachineLearning

OneofthemostcommonAI___7___usedforprocessingbigdataismachinelearning,a

self-adaptivealgorithmthatgetsincreasinglybetteranalysisandpatternswithexperienceorwith

newlyaddeddata.

IfadigitalpaymentscompanywantedtodetecttheoccuiTenceorpotential___8___inits

system,itcouldemploymachinelearningtoolsforthispurpose.Thecomputationalalgorithm

builtintoacomputermodelwillprocessall___9___happeningonthedigitalplatform,find

patternsinthedataset,andpointoutanyanomalydetectedbythepattern.

Deeplearningutilizesahierarchicallevelofartificialneuralnetworkstocarryoutthe

processofmachinelearning.Theartificialneuralnetworksarebuiltlikethehumanbrain,with

neuron___10___connectedtogetherlikeaweb.Whiletraditionalprogramsbuildanalysiswith

datainalinearway,thehierarchicalfunctionofdeeplearningsystemsenablesmachinesto

processdatawithanonlinearapproach.

填空題2

供選擇的答案:

storedresolutionmatchlookunlock

databasephotographeyesreturn,identifying

FaceRecognition

Facerecognitionsystemsusecomputeralgorithmstopickoutspecific,distinctivedetails

aboutaperson'sface.Thesedetails,suchasdistancebetweenthe___1___orshapeofthechin,

arethenconvertedintoamathematicalrepresentationandcomparedtodataonotherfaces

collectedinafacerecognitiondatabase.Thedataaboutaparticularfaceisoftencalledaface

templateandisdistinctfroma___2___becauseit'sdesignedtoonlyincludecertaindetailsthat

canbeusedtodistinguishonefacefromanother.

Somefacerecognitionsystems,insteadofpositively___3___anunknownperson,are

designedtocalculateaprobabilitymatchscorebetweentheunknownpersonandspecificface

templates___4___inthedatabase.Thesesystemswillofferupseveralpotentialmatches,ranked

inorderoflikelihoodofcorrectidentification,insteadofjustreturningasingleresult.

Facerecognitionsystemsvaryintheirabilitytoidentifypeopleunderchallengingconditions

suchaspoorlighting,lowqualityimage___5___,andsuboptimalangleofview(suchasina

photographtakenfromabovelookingdownonanunknownperson).

Whenitcomestoenors,therearetwokeyconceptstounderstand:

A<6falsenegative“iswhenthefacerecognitionsystemfailsto___6___matchaperson's

facetoanimagethatis,infact,containedinadatabase.Inotherwords,thesystemwill

erroneously___7___zeroresultsinresponsetoaquery.

A“falsepositive“iswhenthefacerecognitionsystemdoesmatchaperson'sfacetoan

imageina___8___,butthatmatchisactuallyincorrect.Thisiswhenapoliceofficersubmitsan

imageof"Joe,"butthesystemerroneouslytellstheofficerthatthephotoisof"Jack.”

Whenresearchingafacerecognitionsystem,itisimportantto___9___closelyatthe"false

positive“rateandthe“falsenegative^^rate,sincethereisalmostalwaysatrade-off.Forexample,

ifyouareusingfacerecognitionto___10___yourphone,itisbetterifthesystemfailstoidentify

youafewtimes(falsenegative)thanitisforthesystemtomisidentifyotherpeopleasyouand

letsthosepeopleunlockyourphone(falsepositive).Iftheresultofamisidentificationisthatan

innocentpersongoestojail(likeamisidentificationinamugshotdatabase),thenthesystem

shouldbedesignedtohaveasfewfalsepositivesaspossible.

六、將下面兩篇短文翻譯成中文(每小題10分，共20分)

短文1

DifferencesbetweenStrongAIandWeakAI

1.Meaning

StrongAIisatheoreticalformofartificialintelligencewhichsupportstheviewthat

machinescanreallydevelophumanintelligenceandconsciousnessinthesamewaythatahuman

inconscious.StrongAIreferstoahypotheticalmachinethatexhibitshumancognitiveabilities.

WeakAI(alsoknownasnarrowAI),ontheotherhand,isaformofartificialintelligencethat

referstotheuseofadvancedalgorithmstoaccomplishspecificproblemsolvingorreasoningtasks

thatdonotencompassthefullrangeofhumancognitiveabilities.

2.Functionality

FunctionsarelimitedinweakAIascomparedtostrongAI.WeakAIdoesnotachieve

self-awarenessordemonstrateawiderangeofhumancognitiveabilitiesthatahumanmayhave.

WeakAIreferstosystemsthatareprogrammedtoaccomplishawiderangeproblemsbutoperate

withinapre-determinedorpre-definedrangeoffunctions.StrongAI,ontheotherhand,refersto

machinesthatexhibithumanintelligence.Theideaistodevelopartificialintelligencetothepoint

wherehumaninteractwithmachinesthatareconscious,intelligentanddrivenbyemotionsand

self-awareness.

3.Goal

ThegoalofweakAIistocreateatechnologythatallowsallowsmachinesandcomputersto

toaccomplishspecificproblemsolvingorreasoningtasksatasignificantlyquickerpacethana

humancan.Butitdoesnotnecessarilyincorporateanyrealworldknowledgeabouttheworldof

theproblemthatisbeingsolved.ThegoalofstrongAIistodevelopartificialintelligencetothe

pointwhereitcanbeconsideredtruehumanintelligence.StrongAIisatypeofwhichdoesnot

existyetinitstrueform.

短文2

PatternRecognition

PatternRecognitionisdefinedastheprocessofidentifyingthetrends(globalorlocal)inthe

givenpattern.Apatterncanbedefinedasanythingthatfollowsatrendandexhibitssomekindof

regularity.Therecognitionofpatternscanbedonephysically,mathematicallyorbytheuseof

algorithms.Whenwetalkaboutpatternrecognitioninmachinelearning,itindicatestheuseof

powerfulalgorithmsforidentifyingtheregularitiesinthegivendata.Patternrecognitioniswidely

usedinthenewagetechnicaldomainslikecomputervision,speechrecognition,facerecognition,

etc.

Therearetwotypesofpatternrecognitionalgorithmsinmachinelearning.

1.SupervisedAlgorithms

Thepatternrecognitioninasupervisedapproachiscalledclassification.Thesealgorithms

useatwo-stagemethodologyforidentifyingthepatterns.Thefirststageisthedevelopment/

constructionofthemodelandthesecondstageinvolvesthepredictionforneworunseenobjects.

Thekeyfeaturesinvolvingthisconceptarelistedbelow.

?Classifythegivendataintotwosets-trainingsetandtestingset.

?TrainthemodelusingasuitablemachinelearningalgorithmsuchasSVM(SupportVector

Machines),decisiontrees,randomforest,etc.

?Themodelistrainedonthetrainingsetandtestedonthetestingset.

?Theperformanceofthemodelisevaluatedbasedoncorrectpredictionsmade.

2.UnsupervisedAlgorithms

Incontrasttothesupervisedalgorithmsforpatternmakeuseoftrainingandtestingsets,

thesealgorithmsuseagroupbyapproach.Theyobservethepatternsinthedataandgroupthem

basedonthesimilarityintheirfeaturessuchasdimensiontomakeaprediction.Let'ssaythatwe

haveabasketofdifferentkindsoffruitssuchasapples,oranges,pears,andcherries.Weassume

thatwedonotknowthenamesofthefruits.Wekeepthedataasunlabeled.Now,supposewe

encounterasituationwheresomeonecomesandtellsustoidentifyanewfruitthatwasaddedto

thebasket.Insuchacasewemakeuseofaconceptcalledclustering.

?Clusteringcombinesorgroupsitemshavingthesamefeatures.

?Nopreviousknowledgeisavailableforidentifyinganewitem.

?Theyusemachinelearningalgorithmslikehierarchicalandk-mansclustering.

,Basedonthefeaturesorpropertiesofthenewobject,itisassignedtoagrouptomakea

prediction.

參考試卷答案

、寫(xiě)出以下單詞的中文意思（每小題0.5分，共10分）

1accuracyn.精確（性），準(zhǔn)確（性）IIcustomizevt.定制，定做；用戶(hù)化

2actuatorn.執(zhí)行器12definitionn.定義

3adjustV.調(diào)整，調(diào)節(jié)；適應(yīng)；校準(zhǔn)13defuzzificationn.逆模糊化，去模糊化

4agentn.實(shí)體；代理14deploymentn.部署

5algorithmn.算法15effectorn.效應(yīng)器

6analogyn.類(lèi)推16entityn.實(shí)體

7attributen.屬性；性質(zhì)；特征17extractV.提取，提煉

8backtrackvi.回溯18feedbackn反饋

9blockchainn.區(qū)塊鏈19finiteadj.有限的；限定的

1020n.構(gòu)架；框架；（體系的）

clusterv.聚集n.團(tuán)，群，簇framework

結(jié)構(gòu)

二、根據(jù)給出的中文意思，寫(xiě)出英文單詞（每小題0.5分，共10分）

1V.收集，搜集gather11n.神經(jīng)元；神經(jīng)細(xì)胞neuron

2adj.嵌入的，內(nèi)置的inbuilt12n.節(jié)點(diǎn)node

3n.指示器；指標(biāo)indicator13V.運(yùn)轉(zhuǎn)；操作operate

4n.基礎(chǔ)設(shè)施，基礎(chǔ)架構(gòu)infrastructure14n.模式pattern

5v.合并：集成integrate15V.察覺(jué)，發(fā)覺(jué)perceive

6n.解釋器，解釋程序interpreter16n.前提premise

7n.迭代；循環(huán)iteration17adj.程序的；過(guò)程的procedural

8n.庫(kù)library18n.回歸regression

919adj.健壯的，強(qiáng)健的；

n.元數(shù)據(jù)metadatarobust

結(jié)實(shí)的

10V.監(jiān)視；控制；監(jiān)測(cè)monitor20V.篩選screen

三、根據(jù)給出的短語(yǔ)，寫(xiě)出中文意思（每小題1分，共10分）

1dataobject數(shù)據(jù)對(duì)象

2cybersecurity網(wǎng)絡(luò)安全

3smartmanufacturing智能制造

4clusteredsystem集群系統(tǒng)

5datavisualization數(shù)據(jù)可視化

6opensource開(kāi)源

7analyzetext分析文本

8cloudcomputing云計(jì)算

9computationpower計(jì)算能力

10objectrecognition物體識(shí)別

四、根據(jù)給出的中文意思，寫(xiě)出英文短語(yǔ)（每小題1分，共10分）

1數(shù)據(jù)結(jié)構(gòu)datastructure

2決策樹(shù)decisiontree

3演繹推理deductivereasoning

4貪婪最佳優(yōu)先搜索greedybest-firstsearch

5隱藏模式，隱含模式hiddenpattern

6知識(shí)挖掘knowledgemining

7邏輯推理logicalreasoning

8預(yù)測(cè)性維護(hù)predictivemaintenance

9搜索引擎searchengine

10文本挖掘技術(shù)textminingtechnique

五、寫(xiě)出以下縮略語(yǔ)的完整形式和中文意思（每小題1分，共10分）

縮略語(yǔ)完整形式中文意思

1ANNArtificialNeuralNetwork人工神經(jīng)網(wǎng)絡(luò)

2ARAugmentedReality增強(qiáng)現(xiàn)實(shí)

3BFSBreadth-FirstSearch寬度優(yōu)先搜索

4CVComputerVision計(jì)算機(jī)視覺(jué)

5DFSDepth-FirstSearch深度優(yōu)先搜索

6ESExpertSystem專(zhuān)家系統(tǒng)

7IAIntelligentAgent智能體

8KNNK-NearestNeighborK最近鄰算法

9NLPNaturalLanguageProcessing自然語(yǔ)言處理

10VRVirtualReality虛擬現(xiàn)實(shí)

六、閱讀短文，回答問(wèn)題（每小題2分，共10分）

I.Anartificialneuralnetwork(ANN)isthepieceofacomputingsystemdesignedtosimulatethe

waythehumanbrainanalyzesandprocessesinformation.Itisthefoundationofartificial

intelligence(AI)andsolvesproblemsthatwouldproveimpossibleordifficultbyhumanor

statisticalstandards.

2.Eachneuronismadeupofacellbodythatisresponsibleforprocessinginformationbycarrying

informationtowards(inputs)andaway(outputs)fromthebrain.

3.The