Collaboration versus communication The Department of Energys Amchitka Island and the Aleut Communit.docx
- 文档编号:27758386
- 上传时间:2023-07-04
- 格式:DOCX
- 页数:45
- 大小:38.42KB
Collaboration versus communication The Department of Energys Amchitka Island and the Aleut Communit.docx
《Collaboration versus communication The Department of Energys Amchitka Island and the Aleut Communit.docx》由会员分享,可在线阅读,更多相关《Collaboration versus communication The Department of Energys Amchitka Island and the Aleut Communit.docx(45页珍藏版)》请在冰豆网上搜索。
CollaborationversuscommunicationTheDepartmentofEnergysAmchitkaIslandandtheAleutCommunit
SignalProcessing,Volume89,Issue1,January2009,Pages1-13
KarlWerner,MagnusJansson
Closepreview | Relatedarticles | Relatedreferenceworkarticles
AbstractAbstract|Figures/TablesFigures/Tables|ReferencesReferences
Abstract
Manyalgorithmsfortransmissioninmultipleinputmultipleoutput(MIMO)communicationsystemsrelyonsecondorderstatisticsofthechannelrealizations.TheproblemofestimatingsuchsecondorderstatisticsofMIMOchannels,basedonlimitedamountsoftrainingdata,istreatedinthisarticle.ItisassumedthattheKroneckermodelholds.ThisimpliesthatthechannelcovarianceistheKroneckerproductofonecovariancematrixthatisassociatedwiththearrayandthescatteringatthetransmitterandonethatisassociatedwiththereceivearrayandthescatteringatthereceiver.Theproposedestimatorusestrainingdatafromanumberofsignalblocks(receivedduringindependentfadesoftheMIMOchannel)tocomputetheestimate.Thisisincontrasttomethodsthatassumethatthechannelrealizationsaredirectlyavailable,orpossibletoestimatealmostwithouterror.Itisalsodemonstratedhowmethodsthatmakeuseofthetrainingdataindirectlyviachannelestimatescanbebiased.
Anestimatorisderivedthatcan,inanasymptoticallyoptimalway,use,notonlythestructureimpliedbytheKroneckerassumption,butalsolinearstructureonthetransmit-andreceivecovariancematrices.Theperformanceoftheproposedestimatorisanalyzedandnumericalsimulationsillustratetheresultsandalsoprovideinsightintothesmallsamplebehaviouroftheproposedmethod.
ArticleOutline
1.Introduction
2.Datamodel
3.Directapproaches
4.Cramér–Raolowerbound
5.Acovariancematchingapproach
5.1.Derivationoftheestimator
5.2.Weightedlowrankapproximation
5.3.Performanceoftheestimator
6.Implementingthecovariancematchingestimator
7.Modelvalidation
8.Numericalresults
8.1.Simulationsetup
8.2.Results
9.Conclusion
AppendixA.ProofofTheorem1
AppendixB.ProofofLemma1
References
Purchase
$31.50
802
ThefirstdecadeoftheNationalDrugAbuseTreatmentClinicalTrialsNetwork:
Bridgingthegapbetweenresearchandpracticetoimprovedrugabusetreatment OriginalResearchArticle
JournalofSubstanceAbuseTreatment,Volume38,Supplement1,June2010,PagesS4-S13
BettyTai,MicheleM.Straus,DavidLiu,StevenSparenborg,RonJackson,DennisMcCarty
Closepreview | Relatedarticles | Relatedreferenceworkarticles
AbstractAbstract|Figures/TablesFigures/Tables|ReferencesReferences
Abstract
TheNationalInstituteonDrugAbuseestablishedtheNationalDrugAbuseTreatmentClinicalTrialsNetwork(CTN)in1999toimprovethequalityofaddictiontreatmentusingscienceasthevehicle.Thenetworkbringsprovidersfromcommunity-baseddrugabusetreatmentprogramsandscientistsfromuniversity-basedresearchcenterstogetherinanalliancethatfostersbidirectionalcommunicationandcollaboration.Collaborationenhancedtherelevanceofresearchtopracticeandfacilitatedthedevelopmentandimplementationofevidence-basedtreatmentsincommunitypracticesettings.TheCTN's20completedtrialstestedpharmacological,behavioral,andintegratedtreatmentinterventionsforadolescentsandadults;morethan11,000individualsparticipatedinthetrials.ThisarticlereviewstherationalefortheCTN,describesthetranslationofitsguidingprinciplesintoresearchendeavors,andanticipatesthefutureevolutionofclinicalresearchwithintheNetwork.
ArticleOutline
1.Introduction
2.CTNinfrastructureandgovernance
2.1.RegionalNodes
2.2.NIDA'sCCTN
2.3.SteeringCommittee
2.4.Specialinterestgroups
3.Influenceofbidirectionalityonprotocoldevelopmentandimplementation
4.The“firstwave”ofresearchstudies
4.1.Buprenorphine
4.2.MIandmotivationalenhancementtherapy
4.3.Motivationalincentives
5.CTNaccomplishments
5.1.Research
5.2.Disseminationandresearchutilization
6.Futuredirections
StatementofConflictofInterest
Acknowledgements
References
Purchase
$31.50
803
Aneuralmodelforthep-medianproblem OriginalResearchArticle
Computers&OperationsResearch,Volume35,Issue2,February2008,Pages404-416
EnriqueDomínguez,JoséMuñoz
Closepreview | Relatedarticles | Relatedreferenceworkarticles
AbstractAbstract|Figures/TablesFigures/Tables|ReferencesReferences
Abstract
ThereexistseveralneuraltechniquesforsolvingNP-hardcombinatorialoptimizationproblems.Atthebeginningofthe1980s,recurrentneuralnetworkswereshowntobeabletosolveoptimizationproblems.Criticismofthisapproachincludesthetendencyofrecurrentneuralnetworkstoproduceinfeasiblesolutionsandpoorlocalminima.Thispaperproposesanewtechniquewhichalwaysprovidesfeasiblesolutionsandremovesthetuningphasesincetheconstraintsareincorporatedintheneuralarchitectureinsteadoftheenergyfunction,thereforethetuningparametersareunnecessary.
Oneofthemostpopularandwell-knownfacilitylocationproblemsisthep-medianproblem,whichconcernsthelocationofpfacilities(ormedians)inordertominimizethetotalweighteddistancebetweenthedemandpointsandthefacilities.Thereexistseveralheuristicstofindoptimalsolutionsoftheproblembasedonthetraditionalformulation.
Inthispaperanewformulationforthep-medianproblembasedontwotypesofdecisionvariablesandwithonlyn+plinearequalityconstraintsispresented,wherenisthenumberofdemandpointsorcustomersandpisthenumberoffacilities(medians).Also,acompetitiverecurrentneuralnetworkisproposedforthisproblem.Theneuralnetworkconsistsoftwolayers(allocationlayerandlocationlayer)with2npprocessunits.Theprocessunitsconstituten+pgroups,whereonlyoneprocessunitpergroupisactiveatthesametimeandprocessunitsofthesamelayerareupdatedinparallel.Moreover,theenergyfunction(objectivefunction)alwaysdecreasesorremainsconstantasthesystemevolvesaccordingtothedynamicalruleproposed.Theeffectivenessandefficiencyofouralgorithmfordifferentproblemsizesareanalyzedincomparisontoconventionalheuristicmethods.Theresultsshowthatourrecurrentneuralnetworkgeneratesgoodsolutionswithareasonablecomputationaleffort.
Scopeandpurpose
Geographicalinformationsystems(GIS)haveoccupiedtheattentionofmanyresearchesinvolvinganumberofacademicfieldsincludinggeography,civilengineering,computerscience,landuseplanning,andenvironmentalsciences.GIScansupportawiderangeofspatialqueriesthatcanbeusedtosupportlocationstudies.ModelapplicationandmodeldevelopmentarethemajorimpactofGISonthefieldoflocationscience.Thesesystemsaredesignedtostore,retrieve,manipulate,analyze,andmapgeographicaldata.GIScanserveasthesourceofinputdataforalocationmodelanditcanalsobeusedtopresentthemodelresults.Forexample,ifap-medianproblemhastobesolved,thenaGISexecutesaheuristicalgorithmthatreadsthedatafromtheGISanditpresentstheresultsinrealtime.
Location–allocationmodelssimultaneouslylocatefacilitiesandallocatedemandpointstothem.Thesemodelsalsoariseinavarietyofpublicandprivatesectorproblems.Thep-medianisthemostwidelyusedlocation–allocationmodel.Thep-medianmodelisNP-hardanditsdatasetbecameverylargeinrealproblem,soheuristicsolutionsarerequired.Asthesizeofthedatasetgrow,thenumberoffeasiblesolutionsgrowsandthequalityofsolutionsandcomputationtimesfromthemostcommonlyusedheuristicaredeteriorated.ThepurposeofthispaperistodevelopaneuralmodeltobeintegratedinGISsoftware.Thus,weproposeanewformulationforthep-medianproblembasedon2npbinaryvariablesandn+pequalitylinearconstraints.Arecurrentneuralmodelisproposedforsolvingthep-medianproblembasedonthisformulationwithoutthedifficultyinselectingappropriatetuningparameters,sincetheseparametersareavoided.Moreover,annp-parallelalgorithmhasbeendevelopedbasedonthisformulation.Thecentralpropertyofthisalgorithmisthattheobjectivefunctionalwaysdecreases(orremainsconstant)asthealgorithmevolvesaccordingtoitsdynamicalrule.Moreover,thisalgorithmcanbealsoimplementedusingopticalorVLSItechnology.
ArticleOutline
1.Introduction
2.Problemformulation
3.Competitiverecurrentneuralmodel
3.1.Architecture
3.2.Computationaldynamics
3.3.Neuralalgorithms
4.Experimentalresults
5.Conclusions
Acknowledgements
FurtherReading
AppendixA.Proofoftheenergydecrease
References
Purchase
$31.50
804
Obstaclesconstrainedgroupmobilitymodelsinevent-drivenwirelessnetworkswithmovablebasestations OriginalResearchArticle
AdHocNetworks,Volume9,Issue3,May2011,Pages400-417
S.Cristaldi,A.Ferro,R.Giugno,G.Pigola,A.Pulvirenti
Closepreview | Relatedarticles | Relatedreferenceworkarticles
AbstractAbstract|Figures/TablesFigures/Tables|ReferencesReferences
Abstract
Inthispaper,weproposeaprotocolfordynamicreconfigurationofad-hocwirelessnetworkswithmovablebasestationsinpresenceofobstacles.Hostsareassignedtobasestationsaccordingtoaprobabilisticthroughputfunctionbasedonboththequalityofthesignalandthebasestationload.Inordertooptimizespacecoverage,basestationsclusterhostsusingadistributedclusteringalgorithm.Obstaclesmayinterferewi
- 配套讲稿:
如PPT文件的首页显示word图标,表示该PPT已包含配套word讲稿。双击word图标可打开word文档。
- 特殊限制:
部分文档作品中含有的国旗、国徽等图片,仅作为作品整体效果示例展示,禁止商用。设计者仅对作品中独创性部分享有著作权。
- 关 键 词:
- Collaboration versus communication The Department of Energys Amchitka Island and Aleut Communit
![提示](https://static.bdocx.com/images/bang_tan.gif)
链接地址:https://www.bdocx.com/doc/27758386.html