




版權(quán)說(shuō)明:本文檔由用戶提供并上傳,收益歸屬內(nèi)容提供方,若內(nèi)容存在侵權(quán),請(qǐng)進(jìn)行舉報(bào)或認(rèn)領(lǐng)
文檔簡(jiǎn)介
LearningfromObservations
(chapter18)Autumn2012Instructor:WangXiaolongHarbinInstituteofTechnology,ShenzhenGraduateSchoolIntelligentComputationResearchCenter(HITSGSICRC)
LearningfromObservations
(chOutlinesLearningagentsInductivelearningDecisiontreelearningMeasuringlearningperformanceOutlinesLearningagentsLearningLearningisessentialforunknownenvironments,i.e.,whendesignerlacksomniscienceLearningisusefulasasystemconstructionmethod,i.e.,exposetheagenttorealityratherthantryingtowriteitdownLearningmodifiestheagent'sdecisionmechanismstoimproveperformanceLearningLearningisessentialLearningagentsLearningagentsLearningelementDesignofalearningelementisaffectedbyWhichcomponentsoftheperformanceelementaretobelearnedWhatfeedbackisavailabletolearnthesecomponentsWhatrepresentationisusedforthecomponentsTypeoffeedback: Supervisedlearning:involveslearningafunctionfromexamplesofitsinputandoutputs.Unsupervisedlearning:involveslearningpatternsintheinputwhennospecificoutputvaluesaresupplied.Reinforcementlearning:learnfromrewards(reinforcement)LearningelementDesignofaleInductivelearningSimplestform:learnafunctionfromexamplesfisthetargetfunctionAnexampleisapair(x,f(x))Problem:findahypothesishsuchthath≈fgivenatraining
setofexamples(Thisisahighlysimplifiedmodelofreallearning:IgnorespriorknowledgeAssumesexamplesaregiven)InductivelearningSimplestforInductivelearningmethodConstruct/adjusth
toagreewithf
ontrainingset (hisconsistentifitagreeswithf
onallexamples)E.g.,curvefitting:InductivelearningmethodConstInductivelearningmethodConstruct/adjusth
toagreewithfontrainingset (hisconsistentifitagreeswithfonallexamples)E.g.,curvefitting:InductivelearningmethodConstInductivelearningmethodConstruct/adjusth
toagreewithfontrainingset (hisconsistentifitagreeswithfonallexamples)E.g.,curvefitting:InductivelearningmethodConstInductivelearningmethodConstruct/adjusth
toagreewithfontrainingset (hisconsistentifitagreeswithfonallexamples)E.g.,curvefitting:InductivelearningmethodConstInductivelearningmethodConstruct/adjusth
toagreewithfontrainingset (hisconsistentifitagreeswithfonallexamples)E.g.,curvefitting:InductivelearningmethodConstInductivelearningmethodConstruct/adjusth
toagreewithfontrainingset (hisconsistentifitagreeswithfonallexamples)E.g.,curvefitting:Ockham’srazor:preferthesimplesthypothesisconsistentwithdata
--InLatin,itmeans“Entitiesarenottobemultipliedbeyondnecessity”InductivelearningmethodConstLearningdecisiontreesProblem:decidewhethertowaitforatableatarestaurant,basedonthefollowingattributes:Alternate:isthereanalternativerestaurantnearby?Bar:isthereacomfortablebarareatowaitin?Fri/Sat:istodayFridayorSaturday?Hungry:arewehungry?Patrons:numberofpeopleintherestaurant(None,Some,Full)Price:pricerange($,$$,$$$)Raining:isitrainingoutside?Reservation:havewemadeareservation?Type:kindofrestaurant(French,Italian,Thai,Burger)WaitEstimate:estimatedwaitingtime(0-10,10-30,30-60,>60)LearningdecisiontreesProblemAttribute-basedrepresentationsExamplesdescribedbyattributevalues(Boolean,discrete,continuous)E.g.,situationswhereIwill/won'twaitforatable:Classificationofexamplesispositive(T)ornegative(F)Attribute-basedrepresentationDecisiontreesOnepossiblerepresentationforhypothesesE.g.,hereisthe“true”treefordecidingwhethertowait:DecisiontreesOnepossiblerepExpressivenessDecisiontreescanexpressanyfunctionoftheinputattributes.E.g.,forBooleanfunctions,truthtablerow→pathtoleaf:Trivially,thereisaconsistentdecisiontreeforanytrainingsetwithonepathtoleafforeachexample(unlessf
nondeterministicinx)butitprobablywon'tgeneralizetonewexamplesPrefertofindmorecompactdecisiontreesExpressivenessDecisiontreescHypothesisspacesHowmanydistinctdecisiontreeswithnBooleanattributes?=numberofBooleanfunctions=numberofdistincttruthtableswith2nrows=22nE.g.,with6Booleanattributes,thereare18,446,744,073,709,551,616treesMoreexpressivehypothesisspaceincreaseschancethattargetfunctioncanbeexpressedincreasesnumberofhypothesesconsistentwithtrainingset
maygetworsepredictionsHypothesisspacesHowmanydistDecisiontreelearningAim:findasmalltreeconsistentwiththetrainingexamplesIdea:(recursively)choose"mostsignificant"attributeasrootof(sub)treeDecisiontreelearningAim:finChoosinganattributeIdea:agoodattributesplitstheexamplesintosubsetsthatare(ideally)"allpositive"or"allnegative"Patrons?isabetterchoiceChoosinganattributeIdea:agUsinginformationtheoryInformationanswersquestionsThemorecluelessIamabouttheanswerinitially,themoreinformationiscontainedintheanswerInformationContent(Entropy):I(P(v1),…,P(vn))=Σi=1-P(vi)log2P(vi)Foratrainingsetcontainingppositiveexamplesandnnegativeexamples:UsinginformationtheoryInformInformationgainAchosenattributeAdividesthetrainingsetEintosubsetsE1,…,EvaccordingtotheirvaluesforA,whereA
hasvdistinctvalues.InformationGain(IG)orreductioninentropyfromtheattributetest:ChoosetheattributewiththelargestIGInformationgainAchosenattriInformationgainForthetrainingset,p=n=6,I(6/12,6/12)=1bitConsidertheattributesPatronsandType(andotherstoo):PatronshasthehighestIGofallattributesandsoischosenbytheDTLalgorithmastherootInformationgainForthetrainiExamplecontd.Decisiontreelearnedfromthe12examples:Substantiallysimplerthan“true”tree---amorecomplexhypothesisisn’tjustifiedbysmallamountofdataExamplecontd.DecisiontreelePerformancemeasurementHowdoweknowthath≈f
?Usetheoremsofcomputational/statisticallearningtheoryTryhonanewtestsetofexamples(usesamedistributionoverexamplespaceastrainingset)Learningcurve=%correctontestsetasafunctionoftrainingsetsizePerformancemeasurementHowdoSummaryLearningneededforunknownenvironments,lazydesignersLearningagent=performanceelement+learningelementForsupervisedlearning,theaimistofindasimplehypothesisapproximatelyconsistentwithtrainingexamplesDecisiontreelearningusinginformationgainLearningperformance=predictionaccuracymeasuredontestsetSummaryLearningneededforunkAssignmentsEx18.3AssignmentsEx18.3LearningfromObservations
(chapter18)Autumn2012Instructor:WangXiaolongHarbinInstituteofTechnology,ShenzhenGraduateSchoolIntelligentComputationResearchCenter(HITSGSICRC)
LearningfromObservations
(chOutlinesLearningagentsInductivelearningDecisiontreelearningMeasuringlearningperformanceOutlinesLearningagentsLearningLearningisessentialforunknownenvironments,i.e.,whendesignerlacksomniscienceLearningisusefulasasystemconstructionmethod,i.e.,exposetheagenttorealityratherthantryingtowriteitdownLearningmodifiestheagent'sdecisionmechanismstoimproveperformanceLearningLearningisessentialLearningagentsLearningagentsLearningelementDesignofalearningelementisaffectedbyWhichcomponentsoftheperformanceelementaretobelearnedWhatfeedbackisavailabletolearnthesecomponentsWhatrepresentationisusedforthecomponentsTypeoffeedback: Supervisedlearning:involveslearningafunctionfromexamplesofitsinputandoutputs.Unsupervisedlearning:involveslearningpatternsintheinputwhennospecificoutputvaluesaresupplied.Reinforcementlearning:learnfromrewards(reinforcement)LearningelementDesignofaleInductivelearningSimplestform:learnafunctionfromexamplesfisthetargetfunctionAnexampleisapair(x,f(x))Problem:findahypothesishsuchthath≈fgivenatraining
setofexamples(Thisisahighlysimplifiedmodelofreallearning:IgnorespriorknowledgeAssumesexamplesaregiven)InductivelearningSimplestforInductivelearningmethodConstruct/adjusth
toagreewithf
ontrainingset (hisconsistentifitagreeswithf
onallexamples)E.g.,curvefitting:InductivelearningmethodConstInductivelearningmethodConstruct/adjusth
toagreewithfontrainingset (hisconsistentifitagreeswithfonallexamples)E.g.,curvefitting:InductivelearningmethodConstInductivelearningmethodConstruct/adjusth
toagreewithfontrainingset (hisconsistentifitagreeswithfonallexamples)E.g.,curvefitting:InductivelearningmethodConstInductivelearningmethodConstruct/adjusth
toagreewithfontrainingset (hisconsistentifitagreeswithfonallexamples)E.g.,curvefitting:InductivelearningmethodConstInductivelearningmethodConstruct/adjusth
toagreewithfontrainingset (hisconsistentifitagreeswithfonallexamples)E.g.,curvefitting:InductivelearningmethodConstInductivelearningmethodConstruct/adjusth
toagreewithfontrainingset (hisconsistentifitagreeswithfonallexamples)E.g.,curvefitting:Ockham’srazor:preferthesimplesthypothesisconsistentwithdata
--InLatin,itmeans“Entitiesarenottobemultipliedbeyondnecessity”InductivelearningmethodConstLearningdecisiontreesProblem:decidewhethertowaitforatableatarestaurant,basedonthefollowingattributes:Alternate:isthereanalternativerestaurantnearby?Bar:isthereacomfortablebarareatowaitin?Fri/Sat:istodayFridayorSaturday?Hungry:arewehungry?Patrons:numberofpeopleintherestaurant(None,Some,Full)Price:pricerange($,$$,$$$)Raining:isitrainingoutside?Reservation:havewemadeareservation?Type:kindofrestaurant(French,Italian,Thai,Burger)WaitEstimate:estimatedwaitingtime(0-10,10-30,30-60,>60)LearningdecisiontreesProblemAttribute-basedrepresentationsExamplesdescribedbyattributevalues(Boolean,discrete,continuous)E.g.,situationswhereIwill/won'twaitforatable:Classificationofexamplesispositive(T)ornegative(F)Attribute-basedrepresentationDecisiontreesOnepossiblerepresentationforhypothesesE.g.,hereisthe“true”treefordecidingwhethertowait:DecisiontreesOnepossiblerepExpressivenessDecisiontreescanexpressanyfunctionoftheinputattributes.E.g.,forBooleanfunctions,truthtablerow→pathtoleaf:Trivially,thereisaconsistentdecisiontreeforanytrainingsetwithonepathtoleafforeachexample(unlessf
nondeterministicinx)butitprobablywon'tgeneralizetonewexamplesPrefertofindmorecompactdecisiontreesExpressivenessDecisiontreescHypothesisspacesHowmanydistinctdecisiontreeswithnBooleanattributes?=numberofBooleanfunctions=numberofdistincttruthtableswith2nrows=22nE.g.,with6Booleanattributes,thereare18,446,744,073,709,551,616treesMoreexpressivehypothesisspaceincreaseschancethattargetfunctioncanbeexpressedincreasesnumberofhypothesesconsistentwithtrainingset
maygetworsepredictionsHypothesisspacesHowmanydistDecisiontreelearningAim:findasmalltreeconsistentwiththetrainingexamplesIdea:(recursively)choose"mostsignificant"attributeasrootof(sub)treeDecisiontreelearningAim:finChoosinganattributeIdea:agoodattributesplitstheexamplesintosubsetsthatare(ideally)"allpositive"or"allnegative"Patrons?isabetterchoiceChoosinganattributeIdea:agUsinginformationtheoryInformationanswersquestionsThemorecluelessIamabouttheanswerinitially,themoreinformationiscontainedintheanswerInformationContent(Entropy):I(P(v1),…,P(vn))=Σi=1-P(vi)log2P(vi)Foratrainingsetcontainingppositiveexamplesandnnegativeexamples:UsinginformationtheoryInformInformationgainAchosenattributeAdividesthetrainingsetEintosubsetsE1,…,Evaccordingto
溫馨提示
- 1. 本站所有資源如無(wú)特殊說(shuō)明,都需要本地電腦安裝OFFICE2007和PDF閱讀器。圖紙軟件為CAD,CAXA,PROE,UG,SolidWorks等.壓縮文件請(qǐng)下載最新的WinRAR軟件解壓。
- 2. 本站的文檔不包含任何第三方提供的附件圖紙等,如果需要附件,請(qǐng)聯(lián)系上傳者。文件的所有權(quán)益歸上傳用戶所有。
- 3. 本站RAR壓縮包中若帶圖紙,網(wǎng)頁(yè)內(nèi)容里面會(huì)有圖紙預(yù)覽,若沒(méi)有圖紙預(yù)覽就沒(méi)有圖紙。
- 4. 未經(jīng)權(quán)益所有人同意不得將文件中的內(nèi)容挪作商業(yè)或盈利用途。
- 5. 人人文庫(kù)網(wǎng)僅提供信息存儲(chǔ)空間,僅對(duì)用戶上傳內(nèi)容的表現(xiàn)方式做保護(hù)處理,對(duì)用戶上傳分享的文檔內(nèi)容本身不做任何修改或編輯,并不能對(duì)任何下載內(nèi)容負(fù)責(zé)。
- 6. 下載文件中如有侵權(quán)或不適當(dāng)內(nèi)容,請(qǐng)與我們聯(lián)系,我們立即糾正。
- 7. 本站不保證下載資源的準(zhǔn)確性、安全性和完整性, 同時(shí)也不承擔(dān)用戶因使用這些下載資源對(duì)自己和他人造成任何形式的傷害或損失。
最新文檔
- 襪子超市企業(yè)ESG實(shí)踐與創(chuàng)新戰(zhàn)略研究報(bào)告
- 智能煎烤機(jī)行業(yè)深度調(diào)研及發(fā)展戰(zhàn)略咨詢報(bào)告
- 機(jī)場(chǎng)企業(yè)縣域市場(chǎng)拓展與下沉戰(zhàn)略研究報(bào)告
- 機(jī)器人食品異物剔除系統(tǒng)行業(yè)跨境出海戰(zhàn)略研究報(bào)告
- 餃子店企業(yè)數(shù)字化轉(zhuǎn)型與智慧升級(jí)戰(zhàn)略研究報(bào)告
- 食品用乙?;前匪徕浧髽I(yè)數(shù)字化轉(zhuǎn)型與智慧升級(jí)戰(zhàn)略研究報(bào)告
- 2025年生物質(zhì)壓縮成型設(shè)備項(xiàng)目合作計(jì)劃書(shū)
- 二零二五年度生物科技企業(yè)新員工試崗生物安全協(xié)議
- 二零二五年度生態(tài)農(nóng)業(yè)合伙經(jīng)營(yíng)股權(quán)協(xié)議書(shū)
- 2025年度立體車庫(kù)建設(shè)與運(yùn)營(yíng)一體化項(xiàng)目合同
- 2025年遼寧現(xiàn)代服務(wù)職業(yè)技術(shù)學(xué)院?jiǎn)握新殬I(yè)技能測(cè)試題庫(kù)(含答案)
- 高考模擬作文“中國(guó)游”“city不city”導(dǎo)寫及范文
- 福建省福州市2024-2025學(xué)年九年級(jí)上學(xué)期期末語(yǔ)文試題(解析版)
- 2025年江西電力職業(yè)技術(shù)學(xué)院高職單招職業(yè)適應(yīng)性測(cè)試近5年??及鎱⒖碱}庫(kù)含答案解析
- 2025年月度工作日歷含農(nóng)歷節(jié)假日電子表格版
- 部編版六年級(jí)下冊(cè)道德與法治全冊(cè)教案教學(xué)設(shè)計(jì)
- 物流無(wú)人機(jī)垂直起降場(chǎng)選址與建設(shè)規(guī)范
- 第四紀(jì)地質(zhì)與環(huán)境:第十一章 第四紀(jì)氣候變遷及其動(dòng)力機(jī)制
- 小學(xué)生心理健康講座-(精)
- 蝴蝶豌豆花(課堂PPT)
- 口腔修復(fù)學(xué)-第七章-牙列缺失的全口義齒修復(fù)
評(píng)論
0/150
提交評(píng)論