模擬國際會議PPT

一、基本內(nèi)容

標題頁、目錄頁、章節(jié)內(nèi)容、聲明、參考文獻、致謝

其中，章節(jié)內(nèi)容通常包括主題介紹、實驗或者計算過程、結(jié)果、結(jié)論或總結(jié)二、PPT制作步驟

1)確定章節(jié)內(nèi)容，對各部分內(nèi)容進行邏輯性分析和重要性排序2)PPT初步成型3)PPT詳細設(shè)計4)檢查完善三、設(shè)計原則

目的明確、思路清晰、邏輯性強

文字、表格、圖表合理搭配，并善于使用結(jié)構(gòu)圖簡潔大方、有較好的視覺效果四、設(shè)計內(nèi)容版式設(shè)計模板設(shè)計配色設(shè)計動畫設(shè)計切換設(shè)計效果設(shè)計說明：

1)PPT是輔助說明的工具，使表達內(nèi)容達到易于接受、賞心悅目的效果。2)PPT制作熟能生巧，注意搜集好的設(shè)計和素材，制作時信手拈來。

3)PPT的使用效果與演講者的表達技巧密切相關(guān)，演講者應(yīng)該以飽滿的熱情，盡力將自己

熟知的內(nèi)容分享給觀眾。

擴展閱讀：模擬國際會議演講稿

Recsplorer:RecommendationAlgorithmsBasedonPrecedenceMining

1.Introduction

Thankyouverymuch,Dr.Li,foryourkindintroduction.Ladiesandgentlemen,Goodmorning!Iamhonoredtohavebeeninvitedtospeakatthisconference.BeforeIstartmyspeech,letmeaskaquestion.Doyouthinkrecomemdationsfromothersareusefulforyourinternetshopping?Thankyou.Itisobviousthatrecommendationsplayanimportantroleinourdailyconsumptiondecisions.

Today,mytopicisaboutRecommendationAlgorithmsBasedonPrecedenceMining.Iwanttoshareourinterestingresearchresultonrecommendationalgorithmswithyou.Thecontentofthispresentationisdividedinto5parts:insession1,Iwillintruducethetradictionalrecommendationandournewstrategy;insession2,IwillgivetheformaldefinitionofPrecedenceMining;insession3,Iwilltalkaboutthenovelrecommendationalgorithms;experimentalresultwillbeshowedinsession4;andfinally,Iwillmakeaconclusion.

2.Body

Session1:Introduction

Thepictureonthisslideisaninstanceofrecommemdationapplicationonamazon.

Recommendersystemsprovideadviceonproducts,movies,webpages,andmanyothertopics,andhavebecomepopularinmanysites,suchasAmazon.Manysystemsusecollaborativefilteringmethods.ThemainprocessofCFisorganizedasfollow:first,identifyuserssimilartotargetuser;second,recommenditemsbasedonthesimilarusers.Unfortunately,theorderofconsumeditemsisneglect.Inourpaper,weconsideranewrecommendationstrategybasedonprecedencepatterns.Thesepatternsmayencompassuserpreferences,encodesomelogicalorderofoptionsandcapturehowinterestsevolve.

Precedenceminingmodelestimatetheprobabilityofuserfutureconsumptionbasedonpastbehavior.Andtheseprobabilitiesareusedtomakerecommendations.Throughourexperiment,precedenceminingcansignificantlyimproverecommendationperformance.Futhermore,itdoesnotsufferfromthesparsityofratingsproblemandexploitpatternsacrossallusers,notjustsimilarusers.

Thisslidedemonstratesthedifferencesbetweencollaborativefilteringandprecedencemining.Supposethatthescenarioisaboutcourseselection.Eachquarter/semesterastudentchoosesacourse,andratesitfrom1to5.Figurea)showsfivetranscripts,atranscriptmeansalistofcourse.Uisourtargetstudentwhoneedrecommendations.Figureb)illustrateshowCFwork.Assumesimilarusersshareatleasttwocommoncoursesandhavesimilarrating,thenu3andu4aresimilartou,andtheircommoncoursehwillbearecommendationtou.Figurec)presentshowprecedenceminingwork.Forthisexample,weconsiderpatternswhereonecoursefollowsanother.Supposepatternsoccouratleasttwotranscripsarerecognizedassignificant,then(a,d),(e,f)and(g,h)arefoundout.Andd,h,andfarerecommendationtouwhohastakena,gande.

NowIwillaprobabilisticframeworktosolvetheprecedenceminingproblems.Ourtargetuserhasselectedcoursea,wewanttocomputetheprobabilitycoursexwillfollow,i.e.,Pr[x|a].

howerve,whatwereallyneedtocalculateisPr[x|aX]ratherthanPr[x|a].Becauseinourcontext,wearedecidingifxisagoodrecommendationforthetargetuserthathastakena.Thusweknowthatourtargetuser’stranscriptdoesnothavexbeforea.Forinstance,thetranscriptno.5willbeomitted.Inmorecommonsituation,ourtargetuserhastakenalistofcourses,T={a,b,c,…}not

justa.Thus,whatreallyneedisPr[x|TX].Thequestionishowtofigureoutthisprobability.Iwillansweritlater.

Session2:PrecedenceMining

WeconsiderasetDofdistinctcourses.Weuselowercaseletters(e.g.,a,b,…)torefertocoursesinD.AtranscriptTisasequenceofcourses,e.g.,a->b->c->d.ThenthedefinitionofTop-kRecommendationProblemisasfollows.GivenasettranscriptsoverDfornusers,theextratranscriptTofatargetuser,andadesirednumberofrecommendationsk,ourgoalisto:

1.Assignascorescore(x)(between0and1)toeverycoursex∈Dthatreflectshowlikelyitisthetargetstudentwillbeinterestedintakingx.Ifx∈T,thenscore(x)=0.

2.Usingthescorefunction,selectthetopkcoursestorecommendtothetargetuser.Tocomputescores,weproposetousethefollowingstatistics,wherex,y∈D:f(x):thenumberoftranscriptsthatcontainx.

g(x;y):thenumberoftranscriptsinwhichxprecedescoursey.

Thisslideshowsthecalculationresultoff(x)andg(x,y).Forexample,fromthetable,weknowthatf(a)is10andg(a,c)is3.

WeproposeaprecedenceminingmodeltosolvetheTop-kRecommendationProblem.Hereare

somenotation:xy,whichwehavememtionedinsession1,referstotranscriptwherexoccurswithoutaprecedingy;xyreferstotranscriptwherexoccurswithoutyfollowingit.Weusequantitiesf(x)andg(x,y)tocompteprobabilitiesthatencodetheprecedenceinformation.Forinstance,fromformular1to7.Iwouldnottellthedetailofallformulars.Wejustpayattentionto

formular5,notethatthisquantityaboveisthesameas:Pr[xy|yx]whichwillbeusedtocomputescore(x).

Asweknow,thetargetuserusuallyhastakenalistofcoursesratherthanacourse,soweneedto

extentourprobabilitycalculationformulars.Forexample,supposeT={a,b},Pr[xT]theprobabilityxoccurswithouteitheranaorbprecedingit;Pr[xT]theprobabilityxoccurswithouteitheranaorbfollowingit.Thisprobabilitycanbecalculatedexactly.Sohowtocalculateit?

Session3:RecommendationAlgorithms

Let’sreviewsession2.Themaingoaloftherecommendationalgorithmsistocalculatethescore(x),andthenselectthetopkcoursesbasedonthesescores.TraditionalrecommendationalgorithmscomputearecommendationscoreforacoursexinDonlybasedonitsfrequencyofoccurence.Itdoesnottakeintoaccountthecoursestakenbythetargetuser.

OurrecommendationalgorithmscalledSingleMCconquertheshortcomingofthetraditionalones.Itcomputesthescore(x)usingtheformular5.Thedetailisasfollows:astudentwithatranscripToftakencourses,forthecoursey∈T,ifyandxappeartogetherintranscriptssatisfiesthe

thresholdθ,thencomputethePr[xy|yx],reflectingthelikelihoodthestudentwilltakecoursex

andignoringtheeffectoftheothercoursesinT;finallythemaximumofPr[xy|yx]ischoosenasthescore(x).

Hereisthecalculationformularofscore(x)ofSignleMC.Forexample,withthehigerscore,dwillberecommended.

AnothernewrecommendationalgorithmnamedJointProbabilitiesalgorithm,JointPforshort,isproposed.UnlikeSingleMC,JointPtakesintoaccountthecompletesetofcoursesinatranscript.Informular12,wecannotcomputeitsquantityexactly,Rememberthisproblemwehavementioned.Oursolutionistouseapproximations.Thisslideisaboutthefirstapproximatingformular.Andthisthesecondapproximatingformular.

ThesystemiscourseRand,anddatasetforexperimentcontains7,500transcripts.

Thisslideshowsthenewrecommendationalgoritmswithblackcolorandthetraditionaloneswithbluecolor.

Thechartonthisslideindicatesournewrecommendationalgorithmsbeatthetraditionalonesinprecision,becausetheformeronesexploitpatternsacrossallusers,whilethelatteronesjustusethesimilarusers.

Thechartonthisslidepointsoutournewrecommendationalgorithmsalsobeatthetraditionalonesincoverageforthesamereason.

Session5:ConclusionandSummary

Inconclusion,weproposedanovelprecedenceminingmodel,developedaprobabilisticframeworkformakingrecommendationsandimplementedasuiteofrecommendationalgorithmsthatusetheprecedenceinformation.Experimentalresultshowsthatournewalgorithmsperformbetterthanthetraditionalones,andourrecommendationsystemcanbeeasilygeneralizedtootherscenarios,suchaspurchasesofbooks,DVDsandelectronicequitment.

Tosumup,first,Iintroducedthemotivationandtheoutlineofwork;second,Igavethedefinitionofprecedenceminingmodel;third,Idescribedsomenewrecommendationalgorithmsusingprecedenceinformation;forth,Ishowedourexperimentalresultstocomparethenewalgorithmswithtraditionalones.Finally,Imadeaconclusionofourwork..

That’sall.Thankyou!Arethereanyquestions?

友情提示：本文中關(guān)于《模擬國際會議PPT》給出的范例僅供您參考拓展思維使用，模擬國際會議PPT：該篇文章建議您自主創(chuàng)作。

來源：網(wǎng)絡(luò)整理 免責(zé)聲明：本文僅限學(xué)習(xí)分享，如產(chǎn)生版權(quán)問題，請聯(lián)系我們及時刪除。

《模擬國際會議PPT》由互聯(lián)網(wǎng)用戶整理提供,轉(zhuǎn)載分享請保留原作者信息,謝謝!
鏈接地址：http://www.weilaioem.com/gongwen/668244.html

• 上一篇：如何有效地開展幼兒園區(qū)角活動
• 下一篇：幼兒園區(qū)角活動的指導(dǎo)研究