development of success measures for CHKMSWord文档格式.docx
《development of success measures for CHKMSWord文档格式.docx》由会员分享,可在线阅读,更多相关《development of success measures for CHKMSWord文档格式.docx(61页珍藏版)》请在冰豆网上搜索。
Relatedreferenceworkarticles
Purchase
$37.95
731
AudiosteganalysiswithHausdorffdistancehigherorderstatisticsusingarulebaseddecisiontreeparadigm
ExpertSystemswithApplications,Volume37,Issue12,December2010,Pages7469-7482
S.Geetha,N.Ishwarya,N.Kamaraj
Closepreview
AbstractAbstract|Figures/TablesFigures/Tables|ReferencesReferences
Abstract
Theaimofthispaperistoconstructapracticalforensicsteganalysistoolforaudiosignalsthatcanproperlyanalyzethestatisticsdisturbedbystegoembeddingandclassifythemtoselectedcurrentsteganographicmethods.Theobjectiveofthispaperistoprovethatthechoiceofeffectivestegosensitivefeaturesandaproficientmachinelearningparadigmenhancesthedetectionaccuracyofthesteganalyser.Inthispaperarulebasedapproachwithafamilyofsixdecisiontreeclassifiersviz.,AlternatingDecisionTree,DecisionStump,J48,LogicalModelTree,Naï
veBaye’sTreeandFastDecisionTreelearner,toperformthedetectionofaudiosubliminalchannelisintroduced.InparticularthehigherorderstatisticsextractedfromtheHausdorffdistanceareinvestigatedforanimprovementofthedetectionperformance,ascompetentaudiosteganalyticfeatures.Theevaluationoftheenhancedfeaturespaceandthedecisiontreeparadigm,onadatabasecontaining4800cleanandstegoaudiofilesisperformedforclassicalsteganographicaswellasforwatermarkingalgorithms.WiththisstrategyitisshownhowgeneralforensicapproachcandetectinformationhidingtechniquesinthefieldofcovertcommunicationaswellasforDRMapplications.Forthelattercase,thedetectionofthepresenceofapotentialwatermarkinaspecificfeaturespacecanleadtonewattacksortoabetterdesignofthewatermarkingpattern.
ArticleOutline
1.Motivationandtheapplicationscenarioofaudiosteganalysis
2.Literaturereview
3.Proposedsteganalyser
3.1.Preprocessingoftheaudiosignal
3.2.Featureextraction
3.3.Post-processingoftheresultingfeaturevectors
3.4.Machinelearningparadigm
4.Testscenario
4.1.Testsetsandtestset-up
4.1.1.Informationhidingalgorithmsused
4.1.2.Testfiles
4.1.3.Machinelearningparadigmused
4.2.Testprocedure
4.2.1.Pre-processingoftheaudiodata
4.2.2.Featureextractionformthesignal
4.2.3.Post-processingoftheresultingfeaturevectors
4.2.4.Analysis
4.3.Resultsanddiscussion
4.4.Influenceofembeddingrate
5.Summaryandconclusions
Acknowledgements
References
$41.95
732
Controloflinearmotormachinetoolfeeddrivesforendmilling:
RobustMIMOapproach
Mechatronics,Volume15,Issue10,December2005,Pages1207-1224
ChintaeChoi,Tsu-ChinTsao
Linearmotorsaregettingpromisingforuseashighspeed,highaccuracymachinetoolfeeddrives.Thecuttingforceinthemachiningprocessaredirectlyreflectedtothelinearmotorduetonogearingmechanism.Toachievehighaccuracymachining,thecontrollerforthelinearmotorsystemshouldbedesignedtocompensateforthecuttingforce.
Inthiswork,aMIMOH∞controllerforthelinearmotormachinetoolfeeddriveshasbeendesignedtoreducetrackingerrorsinducedbycuttingforcesforendmilling.Thecontrollerisdesignedusingnormalizedcoprimefactorizationmethodforthedynamicmodelofthelinearmotorsystem.Themodelincludesconstantin-lineandcrosscouplingforcegain,sincethefeedbackcuttingforcecanbeconsideredastheproductoftheconstantgainandthemovingvelocityofeachaxis.
Analysisofthestructuredsingularvalueshowsthatthedesignedcontrollerhasgoodrobustperformancedespitewidevariationsofthecuttingforceandphysicalparameters.ItisdirectlyimplementedonalinearmotorX–YtablewhichismountedonamillingmachinetohavecuttingexperimentsviaaDSPboard.Experimentalresultsverifiedeffectivenessoftheproposedschemetosuppresstheeffectsofthecuttingforceinthehighfeedrate.
1.Introduction
2.Modelingofthelinearmotorsystem
3.H∞loopshapingcontrollerdesign
4.Analysisofrobustness
5.Experimentalresults
6.Conclusions
$31.50
733
Capacitivesensor-basedfluidlevelmeasurementinadynamicenvironmentusingneuralnetwork
EngineeringApplicationsofArtificialIntelligence,Volume23,Issue4,June2010,Pages614-619
EdinTerzic,C.R.Nagarajah,MuhammadAlamgir
Ameasurementsystemhasbeendevelopedusingasingletubecapacitivesensortoaccuratelydeterminethefluidlevelinnon-stationarytanks,namelyautomotivefueltanks.Thesystemdeterminesthefluidlevelinthepresenceofdynamicslosh.Aneuralnetwork-basedapproachisusedtoprocessthesensorsignalandachievesubstantialaccuracycomparedwiththeaveragingmethod,whichisnormallyusedundersuchconditions.Thesensorreadingswereobtainedbyexperimentationcarriedoutundervariousdynamicconditions.Thesensorresponsewasrecordedatvarioussloshfrequenciesandfuelvolumes;
whichwasthenusedtotrainthreedifferentneuralnetworktopologies.FieldtrialswerecarriedouttoobtaintheactualdrivingdataforthepurposeoftestingtheneuralnetworksusingMATLABsoftware.Onestaticneuralnetworktopology,namelyFeed-forwardBackpropagationNeuralNetwork,andtwodynamicneuralnetworktopologies,namelyDistributedTimeDelayNeuralNetworkandNARXNeuralNetwork,havebeeninvestigatedinthiswork.Thedevelopedfluidlevelmeasurementsystemiscapableofdeterminingthefluidlevelinadynamicenvironmentwithamaximumerrorof8.7%byusingthetwodynamicneuralnetworks,and0.11%usingthestaticfeed-forwardbackpropagationneuralnetwork.
2.Neuralnetwork-basedapproach
3.Networktopologies
3.1.Staticfeed-forwardnetwork
3.2.Dynamicneuralnetwork
4.Experimentalsetup
4.1.Trainingdata
4.2.Testdata
5.Simulationresults
6.Summary
7.Futurework
$24.95
734
Standardizationofisletisolationoutcome–Anewautomaticsystemtodeterminepancreaticisletviability
ExpertSystemswithApplications,Volume38,Issue4,April2011,Pages3461-3466
GiadaLaScalia,GiuseppeAiello,CristianaRastellini
Pancreaticislettransplantationisemergingasatherapeuticapproachforpatientsaffectedbydiabetes.Thisapproachconsistsofaminimallyinvasiveprocedurereplacinginsulin-producingcells(pancreaticislets).Thetechniquehasbeenprovensuccessful,butlimitationshavebeenidentified.Oneofthemajorchallengesoftheprocedureisthecountingoftheisolatedpancreaticislets,whichiscurrentlyjeopardizedbysubjectivityandinaccuracy.Determinationoftheaccurateisletnumberisacrucialfactorindeterminingthecorrelationbetweentheisolationproductandclinicaloutcome.Intheproposedstudy,wehavedevelopedsoftwarecapableofobjectivelyevaluatingisletnumbersandotherviabilityvariablesbyimageanalysis.Thissoftwareisbasedonimageprocessingandfeatureextractionalgorithmsforrecognitionoftheareaofinterest.Thisisthefirststeptowardstandardizationoftheisolationoutcomeandpotentialclinicalsuccesspredictability.
2.Theory
3.Methods
3.1.Imageanalysis
3.2.Imageacquisition
3.3.Imageprocessing
4.Featureextraction
5.Resultsanddiscussion
6.Conclusion
Researchhighlights
►AutomaticallydeterminetheamountandthesizeofisletsofLangerhans.►Standardizationofthecountingoftheisolatedpancreaticislets.►Implementationofadditionalindicators.
735
Apixellatedγ-camerabasedonCdTedetectorsclinicalinterestsandperformances
NuclearInstrumentsandMethodsinPhysicsResearchSectionA:
Accelerators,Spectrometers,DetectorsandAssociatedEquipment,Volume448,Issue3,1July2000,Pages537-549
J.Chambron,Y.Arntz,B.Eclancher,Ch.Scheiber,P.Siffert,M.HageHali,R.Regal,A.Kazandjian,V.Prat,S.Thomas,S.Warren,R.Matz,A.Jahnke,M.Karman,A.Pszota,L.Nemeth
Amobilegammacameradedicatedtonuclearcardiology,basedona15
cm×
15
cmdetectionmatrixof2304CdTedetectorelements,2.83
mm×
2.83
2
mm,hasbeendevelopedwithaEuropeanCommunitysupporttoacademicandindustrialresearchcentres.Theintrinsicpropertiesofthesemiconductorcrystals–low-ionisationenergy,high-energyresolution,highattenuationcoefficient–arepotentiallyattractivetoimprovetheγ-cameraperformances.Buttheiruseasγdetectorsformedicalimag