决策层特征融合decisionlevelidentityfusion.ppt

《决策层特征融合decisionlevelidentityfusion.ppt》由会员上传分享，免费在线阅读，更多相关内容在工程资料-天天文库。

1、Decision-LevelIdentityFusionTanXinLab5,SystemEngineeringDept.Contents1.Introduction2.Classicalinference3.Bayesianinference4.Dempster-Shafer’smethod*5.GeneralizedEvidenceProcessing(GEP)Theory6.Heuristicmethodsforidentityfusion7.Implementationandtrade-offsIntroductionDecision-levelfusionSeek

2、stoprocessidentitydeclarationsfrommultiplesensorstoachieveajointdeclarationofidentity.(Featureextraction,identitydeclaration)Data-levelfusionFeature-levelfusionDecision-levelfusion(Datafused)(jointidentitydeclaration)IntroductionSensorASensorBSensorNFeatureExtractionIdentityDeclarationIden

3、tityDeclarationIdentityDeclarationAssociationDecisionLevelFusion–IdentityFusionIntroductionDecision-LevelFusionTechniquesClassicalinferenceBayesianinferenceDempster-Shafer’smethodGeneralizedevidenceprocessingtheoryHeuristicmethodsClassicalinferenceStatisticalinferencetechniquesseektodrawco

4、nclusionsaboutanunderlyingmechanismordistribution,basedonanobservedsampleofdata.Classicalinferencetypicallyassumesanempiricalprobabilitymodel.Empiricalprobabilityassumesthattheobservedfrequencydistributionwillapproximatetheprobabilityasthenumberoftrials.herentrials,occurrenceofktimesTheore

5、ticalbaseClassicalinferenceOnedisadvantageStrictlyspeaking,empiricalprobabilitiesareonlydefinedforrepeatableevents.Classicalinferencemethodsutilizeempiricalprobabilityandhencearenotstrictlyapplicabletononrepeatableevents,unlesssomemodelcanbedevelopedtocomputetherequisiteprobabilities.Class

6、icalinferenceMaintechnique–hypothesistestingDefinetwohypothesis1.Anullhypothesis,H0(原假设)2.Analternativehypothesis,H1(备择假设)Testlogic1.Assumethatthenullhypothesis(H0)istrue;2.ExaminetheconsequencesofH0beingtrueinthesamplingdistributionforstatistic;3.Performahypothesistest,iftheobservationhav

7、eahighprobabilityofbeingobservedifH0istrue,thedeclarethedatadonotcontradictH0.4.Otherwise,declarethatthedatatendtocontradictH0.ClassicalinferenceMaintechnique–hypothesistestingTwoassumptionsarerequired1.anexhaustiveandmutuallyexclusivesetofhypothesiscanbedefin