bayesian network - wikipedia, the free encyclopedia

15
A simple Bayesian network. Rain influences whether the sprinkler is activated, and both rain and the sprinkler influence whether the grass is wet. Bayesian network From Wikipedia, the free encyclopedia A Bayesian network, Bayes network, belief network, Bayes(ian) model or probabilistic directed acyclic graphical model is a probabilistic graphical model (a type of statistical model) that represents a set of random variables and their conditional dependencies via a directed acyclic graph (DAG). For example, a Bayesian network could represent the probabilistic relationships between diseases and symptoms. Given symptoms, the network can be used to compute the probabilities of the presence of various diseases. Formally, Bayesian networks are DAGs whose nodes represent random variables in the Bayesian sense: they may be observable quantities, latent variables, unknown parameters or hypotheses. Edges represent conditional dependencies; nodes that are not connected represent variables that are conditionally independent of each other. Each node is associated with a probability function that takes, as input, a particular set of values for the node's parent variables, and gives (as output) the probability (or probability distribution, if applicable) of the variable represented by the node. For example, if parent nodes represent Boolean variables then the probability function could be represented by a table of entries, one entry for each of the possible combinations of its parents being true or false. Similar ideas may be applied to undirected, and possibly cyclic, graphs; such are called Markov networks. Efficient algorithms exist that perform inference and learning in Bayesian networks. Bayesian networks that model sequences of variables (e.g. speech signals or protein sequences) are called dynamic Bayesian networks. Generalizations of Bayesian networks that can represent and solve decision problems under uncertainty are called influence diagrams. Contents 1 Example 2 Inference and learning 2.1 Inferring unobserved variables 2.2 Parameter learning 2.3 Structure learning 3 Statistical introduction 3.1 Introductory examples 3.2 Restrictions on priors 4 Definitions and concepts 4.1 Factorization definition 4.2 Local Markov property 4.3 Developing Bayesian networks 4.4 Markov blanket 4.4.1 dseparation 4.5 Hierarchical models 4.6 Causal networks 5 Applications 5.1 Software

Upload: lamyanting

Post on 25-Sep-2015

220 views

Category:

Documents


5 download

DESCRIPTION

Bayesian Network

TRANSCRIPT

  • 5/21/2015 BayesiannetworkWikipedia,thefreeencyclopedia

    http://en.wikipedia.org/wiki/Bayesian_network 1/15

    AsimpleBayesiannetwork.Raininfluenceswhetherthesprinklerisactivated,andbothrainandthesprinklerinfluencewhetherthegrassiswet.

    BayesiannetworkFromWikipedia,thefreeencyclopedia

    ABayesiannetwork,Bayesnetwork,beliefnetwork,Bayes(ian)modelorprobabilisticdirectedacyclicgraphicalmodelisaprobabilisticgraphicalmodel(atypeofstatisticalmodel)thatrepresentsasetofrandomvariablesandtheirconditionaldependenciesviaadirectedacyclicgraph(DAG).Forexample,aBayesiannetworkcouldrepresenttheprobabilisticrelationshipsbetweendiseasesandsymptoms.Givensymptoms,thenetworkcanbeusedtocomputetheprobabilitiesofthepresenceofvariousdiseases.

    Formally,BayesiannetworksareDAGswhosenodesrepresentrandomvariablesintheBayesiansense:theymaybeobservablequantities,latentvariables,unknownparametersorhypotheses.Edgesrepresentconditionaldependenciesnodesthatarenotconnectedrepresentvariablesthatareconditionallyindependentofeachother.Eachnodeisassociatedwithaprobabilityfunctionthattakes,asinput,aparticularsetofvaluesforthenode'sparentvariables,andgives(asoutput)theprobability(orprobabilitydistribution,ifapplicable)ofthevariablerepresentedbythenode.Forexample,if parentnodesrepresent Booleanvariablesthentheprobabilityfunctioncouldberepresentedbyatableof entries,oneentryforeachofthe possiblecombinationsofitsparentsbeingtrueorfalse.Similarideasmaybeappliedtoundirected,andpossiblycyclic,graphssucharecalledMarkovnetworks.

    EfficientalgorithmsexistthatperforminferenceandlearninginBayesiannetworks.Bayesiannetworksthatmodelsequencesofvariables(e.g.speechsignalsorproteinsequences)arecalleddynamicBayesiannetworks.GeneralizationsofBayesiannetworksthatcanrepresentandsolvedecisionproblemsunderuncertaintyarecalledinfluencediagrams.

    Contents

    1Example2Inferenceandlearning

    2.1Inferringunobservedvariables2.2Parameterlearning2.3Structurelearning

    3Statisticalintroduction3.1Introductoryexamples3.2Restrictionsonpriors

    4Definitionsandconcepts4.1Factorizationdefinition4.2LocalMarkovproperty4.3DevelopingBayesiannetworks4.4Markovblanket

    4.4.1dseparation4.5Hierarchicalmodels4.6Causalnetworks

    5Applications

    5.1Software

    http://en.wikipedia.org/wiki/Probability_functionhttp://en.wikipedia.org/wiki/Peptide_sequencehttp://en.wikipedia.org/wiki/Conditional_independencehttp://en.wikipedia.org/wiki/Machine_learninghttp://en.wikipedia.org/wiki/Dynamic_Bayesian_networkhttp://en.wikipedia.org/wiki/Latent_variablehttp://en.wikipedia.org/wiki/Boolean_data_typehttp://en.wikipedia.org/wiki/Directed_acyclic_graphhttp://en.wikipedia.org/wiki/Inferencehttp://en.wikipedia.org/wiki/Random_variableshttp://en.wikipedia.org/wiki/Graphical_modelhttp://en.wikipedia.org/wiki/Statistical_modelhttp://en.wikipedia.org/wiki/File:SimpleBayesNetNodes.svghttp://en.wikipedia.org/wiki/Markov_networkhttp://en.wikipedia.org/wiki/Influence_diagramshttp://en.wikipedia.org/wiki/Conditional_independencehttp://en.wikipedia.org/wiki/Glossary_of_graph_theory#Directed_acyclic_graphshttp://en.wikipedia.org/wiki/Speech_recognitionhttp://en.wikipedia.org/wiki/Bayesian_probability
  • 5/21/2015 BayesiannetworkWikipedia,thefreeencyclopedia

    http://en.wikipedia.org/wiki/Bayesian_network 2/15

    AsimpleBayesiannetworkwithconditionalprobabilitytables

    5.1Software6History7Seealso8Notes9References10Furtherreading11Externallinks

    Example

    Supposethattherearetwoeventswhichcouldcausegrasstobewet:eitherthesprinklerisonorit'sraining.Also,supposethattherainhasadirecteffectontheuseofthesprinkler(namelythatwhenitrains,thesprinklerisusuallynotturnedon).ThenthesituationcanbemodeledwithaBayesiannetwork(shown).Allthreevariableshavetwopossiblevalues,T(fortrue)andF(forfalse).

    Thejointprobabilityfunctionis:

    wherethenamesofthevariableshavebeenabbreviatedtoG=Grasswet(yes/no),S=Sprinklerturnedon(yes/no),andR=Raining(yes/no).

    Themodelcananswerquestionslike"Whatistheprobabilitythatitisraining,giventhegrassiswet?"byusingtheconditionalprobabilityformulaandsummingoverallnuisancevariables:

    Usingtheexpansionforthejointprobabilityfunction andtheconditionalprobabilitiesfromtheconditionalprobabilitytables(CPTs)statedinthediagram,onecanevaluateeachterminthesumsinthenumeratoranddenominator.Forexample,

    Thenthenumericalresults(subscriptedbytheassociatedvariablevalues)are

    http://en.wikipedia.org/wiki/Conditional_probability_tablehttp://en.wikipedia.org/wiki/Conditional_probabilityhttp://en.wikipedia.org/wiki/Conditional_probability_tablehttp://en.wikipedia.org/wiki/File:SimpleBayesNet.svghttp://en.wikipedia.org/wiki/Nuisance_variablehttp://en.wikipedia.org/wiki/Joint_probability_distribution
  • 5/21/2015 BayesiannetworkWikipedia,thefreeencyclopedia

    http://en.wikipedia.org/wiki/Bayesian_network 3/15

    If,ontheotherhand,wewishtoansweraninterventionalquestion:"Whatisthelikelihoodthatitwouldrain,giventhatwewetthegrass?"theanswerwouldbegovernedbythepostinterventionjointdistributionfunction obtainedbyremovingthefactor

    fromthepreinterventiondistribution.Asexpected,thelikelihoodofrainisunaffectedbytheaction: .

    If,moreover,wewishtopredicttheimpactofturningthesprinkleron,wehave

    withtheterm removed,showingthattheactionhasaneffectonthegrassbutnotontherain.

    Thesepredictionsmaynotbefeasiblewhensomeofthevariablesareunobserved,asinmostpolicyevaluationproblems.Theeffectoftheaction canstillbepredicted,however,wheneveracriterioncalled"backdoor"issatisfied.[1][2]Itstatesthat,ifasetZofnodescanbeobservedthatdseparates[3](orblocks)allbackdoorpathsfromXtoYthen

    .AbackdoorpathisonethatendswithanarrowintoX.Setsthatsatisfythebackdoorcriterionarecalled"sufficient"or"admissible."Forexample,thesetZ=RisadmissibleforpredictingtheeffectofS=TonG,becauseRdseparatethe(only)backdoorpathSRG.However,ifSisnotobserved,thereisnoothersetthatdseparatesthispathandtheeffectofturningthesprinkleron(S=T)onthegrass(G)cannotbepredictedfrompassiveobservations.WethensaythatP(G|do(S=T))isnot"identified."Thisreflectsthefactthat,lackinginterventionaldata,wecannotdetermineiftheobserveddependencebetweenSandGisduetoacausalconnectionorisspurious(apparentdependencearisingfromacommoncause,R).(seeSimpson'sparadox)

    TodeterminewhetheracausalrelationisidentifiedfromanarbitraryBayesiannetworkwithunobservedvariables,onecanusethethreerulesof"docalculus"[1][4]andtestwhetheralldotermscanberemovedfromtheexpressionofthatrelation,thusconfirmingthatthedesiredquantityisestimablefromfrequencydata.[5]

    UsingaBayesiannetworkcansaveconsiderableamountsofmemory,ifthedependenciesinthejointdistributionaresparse.Forexample,anaivewayofstoringtheconditionalprobabilitiesof10twovaluedvariablesasatablerequiresstoragespacefor values.Ifthelocaldistributionsofnovariabledependsonmorethan3parentvariables,theBayesiannetworkrepresentationonlyneedstostoreatmost values.

    OneadvantageofBayesiannetworksisthatitisintuitivelyeasierforahumantounderstand(asparsesetof)directdependenciesandlocaldistributionsthancompletejointdistributions.

    Inferenceandlearning

    TherearethreemaininferencetasksforBayesiannetworks.

    http://en.wikipedia.org/wiki/Simpson%27s_paradox
  • 5/21/2015 BayesiannetworkWikipedia,thefreeencyclopedia

    http://en.wikipedia.org/wiki/Bayesian_network 4/15

    Inferringunobservedvariables

    BecauseaBayesiannetworkisacompletemodelforthevariablesandtheirrelationships,itcanbeusedtoanswerprobabilisticqueriesaboutthem.Forexample,thenetworkcanbeusedtofindoutupdatedknowledgeofthestateofasubsetofvariableswhenothervariables(theevidencevariables)areobserved.Thisprocessofcomputingtheposteriordistributionofvariablesgivenevidenceiscalledprobabilisticinference.Theposteriorgivesauniversalsufficientstatisticfordetectionapplications,whenonewantstochoosevaluesforthevariablesubsetwhichminimizesomeexpectedlossfunction,forinstancetheprobabilityofdecisionerror.ABayesiannetworkcanthusbeconsideredamechanismforautomaticallyapplyingBayes'theoremtocomplexproblems.

    Themostcommonexactinferencemethodsare:variableelimination,whicheliminates(byintegrationorsummation)thenonobservednonqueryvariablesonebyonebydistributingthesumovertheproductcliquetreepropagation,whichcachesthecomputationsothatmanyvariablescanbequeriedatonetimeandnewevidencecanbepropagatedquicklyandrecursiveconditioningandAND/ORsearch,whichallowforaspacetimetradeoffandmatchtheefficiencyofvariableeliminationwhenenoughspaceisused.Allofthesemethodshavecomplexitythatisexponentialinthenetwork'streewidth.Themostcommonapproximateinferencealgorithmsareimportancesampling,stochasticMCMCsimulation,minibucketelimination,loopybeliefpropagation,generalizedbeliefpropagation,andvariationalmethods.

    Parameterlearning

    InordertofullyspecifytheBayesiannetworkandthusfullyrepresentthejointprobabilitydistribution,itisnecessarytospecifyforeachnodeXtheprobabilitydistributionforXconditionaluponX'sparents.ThedistributionofXconditionaluponitsparentsmayhaveanyform.ItiscommontoworkwithdiscreteorGaussiandistributionssincethatsimplifiescalculations.Sometimesonlyconstraintsonadistributionareknownonecanthenusetheprincipleofmaximumentropytodetermineasingledistribution,theonewiththegreatestentropygiventheconstraints.(Analogously,inthespecificcontextofadynamicBayesiannetwork,onecommonlyspecifiestheconditionaldistributionforthehiddenstate'stemporalevolutiontomaximizetheentropyrateoftheimpliedstochasticprocess.)

    Oftentheseconditionaldistributionsincludeparameterswhichareunknownandmustbeestimatedfromdata,sometimesusingthemaximumlikelihoodapproach.Directmaximizationofthelikelihood(oroftheposteriorprobability)isoftencomplexwhenthereareunobservedvariables.Aclassicalapproachtothisproblemistheexpectationmaximizationalgorithmwhichalternatescomputingexpectedvaluesoftheunobservedvariablesconditionalonobserveddata,withmaximizingthecompletelikelihood(orposterior)assumingthatpreviouslycomputedexpectedvaluesarecorrect.Undermildregularityconditionsthisprocessconvergesonmaximumlikelihood(ormaximumposterior)valuesforparameters.

    AmorefullyBayesianapproachtoparametersistotreatparametersasadditionalunobservedvariablesandtocomputeafullposteriordistributionoverallnodesconditionaluponobserveddata,thentointegrateouttheparameters.Thisapproachcanbeexpensiveandleadtolargedimensionmodels,soinpracticeclassicalparametersettingapproachesaremorecommon.

    Structurelearning

    Inthesimplestcase,aBayesiannetworkisspecifiedbyanexpertandisthenusedtoperforminference.Inotherapplicationsthetaskofdefiningthenetworkistoocomplexforhumans.Inthiscasethenetworkstructureandtheparametersofthelocaldistributionsmustbelearnedfromdata.

    http://en.wikipedia.org/w/index.php?title=Recursive_conditioning&action=edit&redlink=1http://en.wikipedia.org/w/index.php?title=Clique_tree_propagation&action=edit&redlink=1http://en.wikipedia.org/wiki/Loopy_belief_propagationhttp://en.wikipedia.org/wiki/Joint_probability_distributionhttp://en.wikipedia.org/wiki/Treewidthhttp://en.wikipedia.org/wiki/Markov_chain_Monte_Carlohttp://en.wikipedia.org/wiki/Approximate_inferencehttp://en.wikipedia.org/wiki/Posterior_probabilityhttp://en.wikipedia.org/wiki/Space-time_tradeoffhttp://en.wikipedia.org/wiki/Dynamic_Bayesian_networkhttp://en.wikipedia.org/wiki/Expectation-maximization_algorithmhttp://en.wikipedia.org/w/index.php?title=AND/OR_search&action=edit&redlink=1http://en.wikipedia.org/w/index.php?title=Mini-bucket_elimination&action=edit&redlink=1http://en.wikipedia.org/wiki/Maximum_likelihoodhttp://en.wikipedia.org/wiki/Sufficient_statistichttp://en.wikipedia.org/wiki/Bayes%27_theoremhttp://en.wikipedia.org/wiki/Principle_of_maximum_entropyhttp://en.wikipedia.org/wiki/Entropy_ratehttp://en.wikipedia.org/wiki/Variational_Bayeshttp://en.wikipedia.org/wiki/Generalized_belief_propagationhttp://en.wikipedia.org/wiki/Importance_samplinghttp://en.wikipedia.org/wiki/Normal_distributionhttp://en.wikipedia.org/wiki/Information_entropyhttp://en.wikipedia.org/wiki/Variable_elimination
  • 5/21/2015 BayesiannetworkWikipedia,thefreeencyclopedia

    http://en.wikipedia.org/wiki/Bayesian_network 5/15

    AutomaticallylearningthegraphstructureofaBayesiannetworkisachallengepursuedwithinmachinelearning.ThebasicideagoesbacktoarecoveryalgorithmdevelopedbyRebaneandPearl(1987)[6]andrestsonthedistinctionbetweenthethreepossibletypesofadjacenttripletsallowedinadirectedacyclicgraph(DAG):

    1.2.3.

    Type1andtype2representthesamedependencies( and areindependentgiven )andare,therefore,indistinguishable.Type3,however,canbeuniquelyidentified,since and aremarginallyindependentandallotherpairsaredependent.Thus,whiletheskeletons(thegraphsstrippedofarrows)ofthesethreetripletsareidentical,thedirectionalityofthearrowsispartiallyidentifiable.Thesamedistinctionapplieswhen and havecommonparents,exceptthatonemustfirstconditiononthoseparents.Algorithmshavebeendevelopedtosystematicallydeterminetheskeletonoftheunderlyinggraphand,then,orientallarrowswhosedirectionalityisdictatedbytheconditionalindependenciesobserved.[1][7][8][9]

    Analternativemethodofstructurallearningusesoptimizationbasedsearch.Itrequiresascoringfunctionandasearchstrategy.Acommonscoringfunctionisposteriorprobabilityofthestructuregiventhetrainingdata.Thetimerequirementofanexhaustivesearchreturningastructurethatmaximizesthescoreissuperexponentialinthenumberofvariables.Alocalsearchstrategymakesincrementalchangesaimedatimprovingthescoreofthestructure.AglobalsearchalgorithmlikeMarkovchainMonteCarlocanavoidgettingtrappedinlocalminima.Friedmanetal.[10][11]discussusingmutualinformationbetweenvariablesandfindingastructurethatmaximizesthis.Theydothisbyrestrictingtheparentcandidatesettoknodesandexhaustivelysearchingtherein.

    Anothermethodconsistsoffocusingonthesubclassofdecomposablemodels,forwhichtheMLEhaveaclosedform.Itisthenpossibletodiscoveraconsistentstructureforhundredsofvariables.[12]

    ABayesiannetworkcanbeaugmentedwithnodesandedgesusingrulebasedmachinelearningtechniques.Inductivelogicprogrammingcanbeusedtominerulesandcreatenewnodes.[13]Statisticalrelationallearning(SRL)approachesuseascoringfunctionbasedontheBayesnetworkstructuretoguidethestructuralsearchandaugmentthenetwork.[14]AcommonSRLscoringfunctionistheareaundertheROCcurve.

    Statisticalintroduction

    Givendata andparameter ,asimpleBayesiananalysisstartswithapriorprobability(prior)andlikelihood tocomputeaposteriorprobability .

    Oftentheprioron dependsinturnonotherparameters thatarenotmentionedinthelikelihood.So,theprior mustbereplacedbyalikelihood ,andaprior onthenewlyintroducedparameters isrequired,resultinginaposteriorprobability

    ThisisthesimplestexampleofahierarchicalBayesmodel.

    http://en.wikipedia.org/wiki/Statistical_relational_learninghttp://en.wikipedia.org/wiki/Posterior_probabilityhttp://en.wikipedia.org/wiki/Likelihood_functionhttp://en.wikipedia.org/wiki/Scoring_functionhttp://en.wikipedia.org/wiki/Exhaustive_searchhttp://en.wikipedia.org/wiki/Tetrationhttp://en.wikipedia.org/wiki/Prior_probabilityhttp://en.wikipedia.org/wiki/Inductive_logic_programminghttp://en.wikipedia.org/wiki/Mutual_informationhttp://en.wikipedia.org/wiki/ROC_curvehttp://en.wikipedia.org/w/index.php?title=Search_strategy&action=edit&redlink=1http://en.wikipedia.org/wiki/Maximum_likelihood_estimatehttp://en.wikipedia.org/wiki/Markov_chain_Monte_Carlohttp://en.wikipedia.org/wiki/Scoring_functionhttp://en.wikipedia.org/wiki/Bayesian_statisticshttp://en.wikipedia.org/wiki/Machine_learninghttp://en.wikipedia.org/wiki/Maxima_and_minimahttp://en.wikipedia.org/wiki/Machine_learninghttp://en.wikipedia.org/wiki/Posterior_probability
  • 5/21/2015 BayesiannetworkWikipedia,thefreeencyclopedia

    http://en.wikipedia.org/wiki/Bayesian_network 6/15

    Theprocessmayberepeatedforexample,theparameters maydependinturnonadditionalparameters ,whichwillrequiretheirownprior.Eventuallytheprocessmustterminate,withpriorsthatdonotdependonanyotherunmentionedparameters.

    Introductoryexamples

    Supposewehavemeasuredthequantities eachwithnormallydistributederrorsofknownstandarddeviation ,

    Supposeweareinterestedinestimatingthe .Anapproachwouldbetoestimatethe usingamaximumlikelihoodapproachsincetheobservationsareindependent,thelikelihoodfactorizesandthemaximumlikelihoodestimateissimply

    However,ifthequantitiesarerelated,sothatforexamplewemaythinkthattheindividual havethemselvesbeendrawnfromanunderlyingdistribution,thenthisrelationshipdestroystheindependenceandsuggestsamorecomplexmodel,e.g.,

    withimproperpriors flat, flat .When ,thisisanidentifiedmodel(i.e.thereexistsauniquesolutionforthemodel'sparameters),andtheposteriordistributionsoftheindividualwilltendtomove,orshrinkawayfromthemaximumlikelihoodestimatestowardstheircommonmean.ThisshrinkageisatypicalbehaviorinhierarchicalBayesmodels.

    Restrictionsonpriors

    Somecareisneededwhenchoosingpriorsinahierarchicalmodel,particularlyonscalevariablesathigherlevelsofthehierarchysuchasthevariable intheexample.TheusualpriorssuchastheJeffreysprioroftendonotwork,becausetheposteriordistributionwillbeimproper(notnormalizable),andestimatesmadebyminimizingtheexpectedlosswillbeinadmissible.

    Definitionsandconcepts

    ThereareseveralequivalentdefinitionsofaBayesiannetwork.Forallthefollowing,letG=(V,E)beadirectedacyclicgraph(orDAG),andletX=(Xv)vVbeasetofrandomvariablesindexedbyV.

    Factorizationdefinition

    XisaBayesiannetworkwithrespecttoGifitsjointprobabilitydensityfunction(withrespecttoaproductmeasure)canbewrittenasaproductoftheindividualdensityfunctions,conditionalontheirparentvariables:[15]

    http://en.wikipedia.org/wiki/Directed_acyclic_graphhttp://en.wikipedia.org/wiki/Probability_density_functionhttp://en.wikipedia.org/w/index.php?title=Identified_model&action=edit&redlink=1http://en.wikipedia.org/wiki/Maximum_likelihoodhttp://en.wikipedia.org/wiki/Standard_deviationhttp://en.wikipedia.org/wiki/Normal_distributionhttp://en.wikipedia.org/wiki/Loss_function#Expected_losshttp://en.wikipedia.org/wiki/Shrinkage_estimatorhttp://en.wikipedia.org/wiki/Admissible_decision_rulehttp://en.wikipedia.org/wiki/Random_variablehttp://en.wikipedia.org/wiki/Jeffreys_priorhttp://en.wikipedia.org/wiki/Improper_priorhttp://en.wikipedia.org/wiki/Product_measure
  • 5/21/2015 BayesiannetworkWikipedia,thefreeencyclopedia

    http://en.wikipedia.org/wiki/Bayesian_network 7/15

    wherepa(v)isthesetofparentsofv(i.e.thoseverticespointingdirectlytovviaasingleedge).

    Foranysetofrandomvariables,theprobabilityofanymemberofajointdistributioncanbecalculatedfromconditionalprobabilitiesusingthechainrule(givenatopologicalorderingofX)asfollows:[15]

    Comparethiswiththedefinitionabove,whichcanbewrittenas:

    foreach whichisaparentof

    Thedifferencebetweenthetwoexpressionsistheconditionalindependenceofthevariablesfromanyoftheirnondescendants,giventhevaluesoftheirparentvariables.

    LocalMarkovproperty

    XisaBayesiannetworkwithrespecttoGifitsatisfiesthelocalMarkovproperty:eachvariableisconditionallyindependentofitsnondescendantsgivenitsparentvariables:[16]

    wherede(v)isthesetofdescendantsandV\de(v)isthesetofnondescendantsofv.

    Thiscanalsobeexpressedintermssimilartothefirstdefinition,as

    foreach whichisnotadescendantofforeach whichisaparentof

    Notethatthesetofparentsisasubsetofthesetofnondescendantsbecausethegraphisacyclic.

    DevelopingBayesiannetworks

    TodevelopaBayesiannetwork,weoftenfirstdevelopaDAGGsuchthatwebelieveXsatisfiesthelocalMarkovpropertywithrespecttoG.SometimesthisisdonebycreatingacausalDAG.WethenascertaintheconditionalprobabilitydistributionsofeachvariablegivenitsparentsinG.Inmanycases,inparticularinthecasewherethevariablesarediscrete,ifwedefinethejointdistributionofXtobetheproductoftheseconditionaldistributions,thenXisaBayesiannetworkwithrespecttoG.[17]

    Markovblanket

    TheMarkovblanketofanodeisthesetofnodesconsistingofitsparents,itschildren,andanyotherparentsofitschildren.ThissetrendersitindependentoftherestofthenetworkthejointdistributionofthevariablesintheMarkovblanketofanodeissufficientknowledgeforcalculatingthedistributionofthenode.XisaBayesiannetworkwithrespecttoGifeverynodeisconditionallyindependentofallothernodesinthenetwork,givenitsMarkovblanket.[16]

    http://en.wikipedia.org/wiki/Conditional_independencehttp://en.wikipedia.org/wiki/Conditional_independencehttp://en.wikipedia.org/wiki/Chain_rule_(probability)http://en.wikipedia.org/wiki/Markov_blankethttp://en.wikipedia.org/wiki/Joint_distributionhttp://en.wikipedia.org/wiki/Cycle_(graph_theory)http://en.wikipedia.org/wiki/Topological_orderinghttp://en.wikipedia.org/wiki/Markov_blanket
  • 5/21/2015 BayesiannetworkWikipedia,thefreeencyclopedia

    http://en.wikipedia.org/wiki/Bayesian_network 8/15

    dseparation

    Thisdefinitioncanbemademoregeneralbydefiningthe"d"separationoftwonodes,wheredstandsfordirectional.[18][19]LetPbeatrail(thatis,acollectionofedgeswhichislikeapath,buteachofwhoseedgesmayhaveanydirection)fromnodeutov.ThenPissaidtobedseparatedbyasetofnodesZifandonlyif(atleast)oneofthefollowingholds:

    1. Pcontainsachain,umv,suchthatthemiddlenodemisinZ,2. Pcontainsafork,umv,suchthatthemiddlenodemisinZ,or3. Pcontainsaninvertedfork(orcollider),umv,suchthatthemiddlenodemisnotinZand

    nodescendantofmisinZ.

    ThusuandvaresaidtobedseparatedbyZifalltrailsbetweenthemaredseparated.Ifuandvarenotdseparated,theyarecalleddconnected.

    XisaBayesiannetworkwithrespecttoGif,foranytwonodesu,v:

    whereZisasetwhichdseparatesuandv.(TheMarkovblanketistheminimalsetofnodeswhichdseparatesnodevfromallothernodes.)

    Hierarchicalmodels

    ThetermhierarchicalmodelissometimesconsideredaparticulartypeofBayesiannetwork,buthasnoformaldefinition.Sometimesthetermisreservedformodelswiththreeormorelevelsofrandomvariablesothertimes,itisreservedformodelswithlatentvariables.Ingeneral,however,anymoderatelycomplexBayesiannetworkisusuallytermed"hierarchical".

    Causalnetworks

    AlthoughBayesiannetworksareoftenusedtorepresentcausalrelationships,thisneednotbethecase:adirectededgefromutovdoesnotrequirethatXviscausallydependentonXu.ThisisdemonstratedbythefactthatBayesiannetworksonthegraphs:

    areequivalent:thatistheyimposeexactlythesameconditionalindependencerequirements.

    AcausalnetworkisaBayesiannetworkwithanexplicitrequirementthattherelationshipsbecausal.TheadditionalsemanticsofthecausalnetworksspecifythatifanodeXisactivelycausedtobeinagivenstatex(anactionwrittenasdo(X=x)),thentheprobabilitydensityfunctionchangestotheoneofthenetworkobtainedbycuttingthelinksfromtheparentsofXtoX,andsettingXtothecausedvaluex.[1]Usingthesesemantics,onecanpredicttheimpactofexternalinterventionsfromdataobtainedpriortointervention.

    Applications

    Bayesiannetworksareusedformodellingbeliefsincomputationalbiologyandbioinformatics(generegulatorynetworks,proteinstructure,geneexpressionanalysis,[20]learningepistasisfromGWASdatasets[21])medicine,[22]biomonitoring,[23]documentclassification,informationretrieval,[24]semantic

    http://en.wikipedia.org/wiki/Markov_blankethttp://en.wikipedia.org/wiki/Protein_structurehttp://en.wikipedia.org/wiki/Medicinehttp://en.wikipedia.org/wiki/Information_retrievalhttp://en.wikipedia.org/wiki/Computational_biologyhttp://en.wikipedia.org/wiki/Bioinformaticshttp://en.wikipedia.org/wiki/Mathematical_modelhttp://en.wikipedia.org/wiki/Gene_regulatory_networkhttp://en.wikipedia.org/wiki/Causalityhttp://en.wikipedia.org/wiki/Latent_variablehttp://en.wikipedia.org/wiki/Biomonitoringhttp://en.wikipedia.org/wiki/Document_classificationhttp://en.wikipedia.org/wiki/Gene_expressionhttp://en.wikipedia.org/wiki/Semantic_search
  • 5/21/2015 BayesiannetworkWikipedia,thefreeencyclopedia

    http://en.wikipedia.org/wiki/Bayesian_network 9/15

    search,[25]imageprocessing,datafusion,decisionsupportsystems,[26]engineering,sportsbetting,[27][28]

    gaming,law,[29][30][31]studydesign[32]andriskanalysis.[33][34][35]TherearetextsapplyingBayesiannetworkstobioinformatics[36]andfinancialandmarketinginformatics.[37]

    Software

    WinBUGSOpenBUGS(website(http://www.openbugs.net/w/FrontPage)),further(opensource)developmentofWinBUGS.OpenMarkov(http://www.openmarkov.org/),opensourcesoftwareandAPIimplementedinJavaGraphicalModelsToolkit(http://melodi.ee.washington.edu/gmtk)(GMTK)GMTKisanopensource,publiclyavailabletoolkitforrapidlyprototypingstatisticalmodelsusingdynamicgraphicalmodels(DGMs)anddynamicBayesiannetworks(DBNs).GMTKcanbeusedforapplicationsandresearchinspeechandlanguageprocessing,bioinformatics,activityrecognition,andanytimeseriesapplication.JustanotherGibbssampler(JAGS)(website(http://wwwfis.iarc.fr/~martyn/software/jags/))Stan(software)(website(http://mcstan.org/))StanisanopensourcepackageforobtainingBayesianinferenceusingtheNoUTurnsampler,avariantofHamiltonianMonteCarlo.It'ssomewhatlikeBUGS,butwithadifferentlanguageforexpressingmodelsandadifferentsamplerforsamplingfromtheirposteriors.RStanistheRinterfacetoStan.PyMC(http://pymcdevs.github.io/pymc/)PyMCisapythonmodulethatimplementsBayesianstatisticalmodelsandfittingalgorithms,includingMarkovchainMonteCarlo.Itsflexibilityandextensibilitymakeitapplicabletoalargesuiteofproblems.Alongwithcoresamplingfunctionality,PyMCincludesmethodsforsummarizingoutput,plotting,goodnessoffitandconvergencediagnostics.GeNIe&Smile(website(http://genie.sis.pitt.edu/))SMILEisaC++libraryforBNandID,andGeNIeisaGUIforitSamIam(website(http://reasoning.cs.ucla.edu/samiam/)),aJavabasedsystemwithGUIandJavaAPIBayesServer(http://www.BayesServer.com/)UserInterfaceandAPIforBayesiannetworks,includessupportfortimeseriesandsequencesBeliefandDecisionNetworksonAIspace(http://www.aispace.org/bayes/index.shtml)BayesiaLab(http://library.bayesia.com/display/HOME/The+BayesiaLab+Library/)byBayesiaHugin(http://www.hugin.com/)Netica(http://www.norsys.com/netica.html)byNorsysdVelox(http://www.aparasw.com/index.php/en)byAparaSoftwareSystemModeler(http://www.inatas.com)byInatasABUnBBayes(http://sourceforge.net/projects/unbbayes/)byGIAUnB(IntelligenceArtificialGroupUniversityofBrasilia)

    History

    Theterm"Bayesiannetworks"wascoinedbyJudeaPearlin1985toemphasizethreeaspects:[38]

    1. Theoftensubjectivenatureoftheinputinformation.2. TherelianceonBayes'conditioningasthebasisforupdatinginformation.3. Thedistinctionbetweencausalandevidentialmodesofreasoning,whichunderscoresThomas

    Bayes'posthumouslypublishedpaperof1763.[39]

    http://www.norsys.com/netica.htmlhttp://en.wikipedia.org/wiki/Decision_support_systemhttp://www.openbugs.net/w/FrontPagehttp://reasoning.cs.ucla.edu/samiam/http://en.wikipedia.org/wiki/Engineeringhttp://en.wikipedia.org/wiki/Lawhttp://pymc-devs.github.io/pymc/http://en.wikipedia.org/wiki/Thomas_Bayeshttp://en.wikipedia.org/wiki/Stan_(software)http://www.inatas.com/http://www.hugin.com/http://mc-stan.org/http://www.openmarkov.org/http://www.bayesserver.com/http://en.wikipedia.org/wiki/Data_fusionhttp://www.aispace.org/bayes/index.shtmlhttp://en.wikipedia.org/wiki/Judea_Pearlhttp://genie.sis.pitt.edu/http://www.aparasw.com/index.php/enhttp://en.wikipedia.org/wiki/OpenBUGShttp://melodi.ee.washington.edu/gmtkhttp://sourceforge.net/projects/unbbayes/http://library.bayesia.com/display/HOME/The+BayesiaLab+Library/http://en.wikipedia.org/wiki/Just_another_Gibbs_samplerhttp://en.wikipedia.org/wiki/Risk_analysishttp://en.wikipedia.org/wiki/WinBUGShttp://en.wikipedia.org/wiki/Image_processinghttp://www-fis.iarc.fr/~martyn/software/jags/http://en.wikipedia.org/wiki/Semantic_search
  • 5/21/2015 BayesiannetworkWikipedia,thefreeencyclopedia

    http://en.wikipedia.org/wiki/Bayesian_network 10/15

    Inthelate1980sJudeaPearl'stextProbabilisticReasoninginIntelligentSystems[40]andRichardE.Neapolitan'stextProbabilisticReasoninginExpertSystems[41]summarizedthepropertiesofBayesiannetworksandestablishedBayesiannetworksasafieldofstudy.

    InformalvariantsofsuchnetworkswerefirstusedbylegalscholarJohnHenryWigmore,intheformofWigmorecharts,toanalysetrialevidencein1913.[30]:6676Anothervariant,calledpathdiagrams,wasdevelopedbythegeneticistSewallWright[42]andusedinsocialandbehavioralsciences(mostlywithlinearparametricmodels).

    Seealso

    Notes

    ArtificialintelligenceBayes'theoremBayesianinferenceBayesianprobabilityBayesianprogrammingBeliefpropagationCausalloopdiagramChowLiutreeComputationalintelligenceComputationalphylogeneticsDeepbeliefnetworkDempsterShafertheoryaGeneralizationofBayes'theoremDynamicBayesiannetworkExpectationmaximizationalgorithmFactorgraphGraphicalmodelHierarchicaltemporalmemoryInfluencediagramJudeaPearlKalmanfilterMachinelearningMemorypredictionframeworkMixturedistributionMixturemodelNaiveBayesclassifierPathanalysisPolytreeSensorfusionSequencealignmentSpeechrecognitionStructuralequationmodelingSubjectivelogicVariableorderBayesiannetworkWigmorechartWorldview

    1. Pearl,Judea(2000).Causality:Models,Reasoning,andInference.CambridgeUniversityPress.ISBN0521773628.OCLC42291253(https://www.worldcat.org/oclc/42291253).

    http://en.wikipedia.org/wiki/Variable-order_Bayesian_networkhttp://en.wikipedia.org/wiki/Speech_recognitionhttp://en.wikipedia.org/wiki/Dynamic_Bayesian_networkhttp://en.wikipedia.org/wiki/Bayesian_inferencehttp://en.wikipedia.org/wiki/Chow%E2%80%93Liu_treehttp://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://en.wikipedia.org/wiki/Mixture_distributionhttp://en.wikipedia.org/wiki/Special:BookSources/0-521-77362-8http://en.wikipedia.org/wiki/Subjective_logichttp://en.wikipedia.org/wiki/Deep_belief_networkhttp://en.wikipedia.org/wiki/Computational_intelligencehttp://en.wikipedia.org/wiki/Bayesian_probabilityhttp://en.wikipedia.org/wiki/OCLChttp://en.wikipedia.org/wiki/Evidence_(law)http://en.wikipedia.org/wiki/John_Henry_Wigmorehttp://en.wikipedia.org/wiki/Sequence_alignmenthttp://en.wikipedia.org/wiki/World_viewhttp://en.wikipedia.org/wiki/Kalman_filterhttp://en.wikipedia.org/wiki/Naive_Bayes_classifierhttp://en.wikipedia.org/wiki/Hierarchical_temporal_memoryhttp://en.wikipedia.org/wiki/Bayes%27_theoremhttp://en.wikipedia.org/wiki/Legal_scholarhttp://en.wikipedia.org/wiki/Social_scienceshttp://en.wikipedia.org/wiki/Wigmore_charthttp://en.wikipedia.org/wiki/Sensor_fusionhttp://en.wikipedia.org/wiki/Sewall_Wrighthttp://en.wikipedia.org/wiki/Memory-prediction_frameworkhttp://en.wikipedia.org/wiki/Path_analysis_(statistics)http://en.wikipedia.org/wiki/Path_analysis_(statistics)http://en.wikipedia.org/wiki/Artificial_intelligencehttp://en.wikipedia.org/wiki/Computational_phylogeneticshttp://en.wikipedia.org/wiki/Factor_graphhttp://en.wikipedia.org/wiki/Bayesian_programminghttp://en.wikipedia.org/wiki/Polytreehttp://en.wikipedia.org/wiki/Machine_learninghttp://en.wikipedia.org/wiki/Influence_diagramhttp://en.wikipedia.org/wiki/Mixture_modelhttp://en.wikipedia.org/wiki/Judea_Pearlhttp://www.worldcat.org/oclc/42291253http://en.wikipedia.org/wiki/Graphical_modelhttp://en.wikipedia.org/wiki/Belief_propagationhttp://en.wikipedia.org/wiki/Structural_equation_modelinghttp://en.wikipedia.org/wiki/Judea_Pearlhttp://en.wikipedia.org/wiki/Wigmore_charthttp://en.wikipedia.org/wiki/Dempster%E2%80%93Shafer_theoryhttp://en.wikipedia.org/wiki/Cambridge_University_Presshttp://en.wikipedia.org/wiki/Trial_(law)http://en.wikipedia.org/wiki/Behavioral_sciencehttp://en.wikipedia.org/wiki/Expectation%E2%80%93maximization_algorithmhttp://en.wikipedia.org/wiki/Causal_loop_diagram
  • 5/21/2015 BayesiannetworkWikipedia,thefreeencyclopedia

    http://en.wikipedia.org/wiki/Bayesian_network 11/15

    2. "TheBackDoorCriterion"(http://bayes.cs.ucla.edu/BOOK2K/ch33.pdf)(PDF).Retrieved20140918.3. "dSeparationwithoutTears"(http://bayes.cs.ucla.edu/BOOK09/ch1112final.pdf)(PDF).Retrieved

    20140918.4. J.,Pearl(1994)."AProbabilisticCalculusofActions"(http://dl.acm.org/ft_gateway.cfm?

    id=2074452&ftid=1062250&dwn=1&CFID=161588115&CFTOKEN=10243006).InLopezdeMantaras,R.Poole,D.UAI'94ProceedingsoftheTenthinternationalconferenceonUncertaintyinartificialintelligence.SanMateoCA:MorganKaufman.pp.454462.ISBN1558603328.

    5. I.Shpitser,J.Pearl,"IdentificationofConditionalInterventionalDistributions"InR.DechterandT.S.Richardson(Eds.),ProceedingsoftheTwentySecondConferenceonUncertaintyinArtificialIntelligence,437444,Corvallis,OR:AUAIPress,2006.

    6. Rebane,G.andPearl,J.,"TheRecoveryofCausalPolytreesfromStatisticalData,"Proceedings,3rdWorkshoponUncertaintyinAI,(Seattle,WA)pages222228,1987

    7. Spirtes,P.Glymour,C.(1991)."Analgorithmforfastrecoveryofsparsecausalgraphs"(http://repository.cmu.edu/cgi/viewcontent.cgi?article=1316&context=philosophy)(PDF).SocialScienceComputerReview9(1):6272.doi:10.1177/089443939100900106(https://dx.doi.org/10.1177%2F089443939100900106).

    8. Spirtes,PeterGlymour,ClarkN.Scheines,Richard(1993).Causation,Prediction,andSearch(http://books.google.com/books?id=VkawQgAACAAJ)(1sted.).SpringerVerlag.ISBN9780387979793.

    9. Verma,ThomasPearl,Judea(1991)."Equivalenceandsynthesisofcausalmodels".InBonissone,P.Henrion,M.Kanal,L.N.Lemmer,J.F.UAI'90ProceedingsoftheSixthAnnualConferenceonUncertaintyinArtificialIntelligence.Elsevier.pp.255270.ISBN0444892648.

    10. Friedman,NirGeiger,DanGoldszmidt,Moises(November1997)."BayesianNetworkClassifiers"(http://link.springer.com/article/10.1023%2FA%3A1007465528199).MachineLearning29(23):131163.doi:10.1023/A:1007465528199(https://dx.doi.org/10.1023%2FA%3A1007465528199).Retrieved24February2015.

    11. Friedman,NirLinial,MichalNachman,IftachPe'er,Dana(August2000)."UsingBayesianNetworkstoAnalyzeExpressionData"(http://online.liebertpub.com/doi/abs/10.1089/106652700750050961).JournalofComputationalBiology7(34):601620.doi:10.1089/106652700750050961(https://dx.doi.org/10.1089%2F106652700750050961).Retrieved24February2015.

    12. Petitjean,F.Webb,G.I.Nicholson,A.E.(2013).Scalingloglinearanalysistohighdimensionaldata(http://www.tinyclues.eu/Research/Petitjean2013ICDM.pdf)(PDF).InternationalConferenceonDataMining.Dallas,TX,USA:IEEE.

    13. Nassif,HoussamWu,YirongPage,DavidBurnside,Elizabeth(2012)."LogicalDifferentialPredictionBayesNet,ImprovingBreastCancerDiagnosisforOlderWomen"(http://pages.cs.wisc.edu/~hous21/papers/AMIA12.pdf)(PDF).AmericanMedicalInformaticsAssociationSymposium(AMIA'12)(Chicago):13301339.Retrieved18July2014.

    14. Nassif,HoussamKuusisto,FinnBurnside,ElizabethSPage,DavidShavlik,JudeSantosCosta,Vitor(2013)."ScoreAsYouLift(SAYL):AStatisticalRelationalLearningApproachtoUpliftModeling"(http://pages.cs.wisc.edu/~hous21/papers/ECML13.pdf)(PDF).EuropeanConferenceonMachineLearning(ECML'13)(Prague):595611.

    15. Russell&Norvig2003,p.496.16. Russell&Norvig2003,p.499.17. Neapolitan,RichardE.(2004).LearningBayesiannetworks(http://books.google.com/books?

    id=OlMZAQAAIAAJ).PrenticeHall.ISBN9780130125347.18. Geiger,DanVerma,ThomasPearl,Judea(1990)."IdentifyingindependenceinBayesianNetworks"

    (http://ftp.cs.ucla.edu/pub/stat_ser/r116.pdf)(PDF).Networks20:507534.doi:10.1177/089443939100900106(https://dx.doi.org/10.1177%2F089443939100900106).

    19. RichardScheines,Dseparation(http://www.andrew.cmu.edu/user/scheines/tutor/dsep.html)20. Friedman,N.Linial,M.Nachman,I.Pe'er,D.(2000)."UsingBayesianNetworkstoAnalyzeExpression

    Data".JournalofComputationalBiology7(34):601620.doi:10.1089/106652700750050961(https://dx.doi.org/10.1089%2F106652700750050961).PMID11108481(https://www.ncbi.nlm.nih.gov/pubmed/11108481).

    21. Jiang,X.Neapolitan,R.E.Barmada,M.M.Visweswaran,S.(2011)."LearningGeneticEpistasisusingBayesianNetworkScoringCriteria"(http://www.biomedcentral.com/14712105/12/89).BMCBioinformatics12:89.doi:10.1186/147121051289(https://dx.doi.org/10.1186%2F147121051289).PMC3080825(https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3080825).PMID21453508(https://www.ncbi.nlm.nih.gov/pubmed/21453508).

    http://bayes.cs.ucla.edu/BOOK-09/ch11-1-2-final.pdfhttp://www.biomedcentral.com/1471-2105/12/89http://en.wikipedia.org/wiki/Special:BookSources/978-0-387-97979-3http://dx.doi.org/10.1186%2F1471-2105-12-89http://pages.cs.wisc.edu/~hous21/papers/ECML13.pdfhttp://en.wikipedia.org/wiki/Digital_object_identifierhttp://www.ncbi.nlm.nih.gov/pubmed/11108481http://pages.cs.wisc.edu/~hous21/papers/AMIA12.pdfhttp://en.wikipedia.org/wiki/Special:BookSources/1-55860-332-8http://en.wikipedia.org/wiki/Digital_object_identifierhttp://dx.doi.org/10.1089%2F106652700750050961http://dx.doi.org/10.1177%2F089443939100900106http://en.wikipedia.org/wiki/Digital_object_identifierhttp://dx.doi.org/10.1177%2F089443939100900106http://en.wikipedia.org/wiki/PubMed_Identifierhttp://en.wikipedia.org/wiki/PubMed_Centralhttp://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://www.ncbi.nlm.nih.gov/pmc/articles/PMC3080825http://books.google.com/books?id=OlMZAQAAIAAJhttp://en.wikipedia.org/wiki/Digital_object_identifierhttp://bayes.cs.ucla.edu/BOOK-2K/ch3-3.pdfhttp://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://dl.acm.org/ft_gateway.cfm?id=2074452&ftid=1062250&dwn=1&CFID=161588115&CFTOKEN=10243006http://repository.cmu.edu/cgi/viewcontent.cgi?article=1316&context=philosophyhttp://dx.doi.org/10.1023%2FA%3A1007465528199http://en.wikipedia.org/wiki/Digital_object_identifierhttp://www.andrew.cmu.edu/user/scheines/tutor/d-sep.htmlhttp://www.ncbi.nlm.nih.gov/pubmed/21453508http://dx.doi.org/10.1089%2F106652700750050961http://www.tiny-clues.eu/Research/Petitjean2013-ICDM.pdfhttp://en.wikipedia.org/wiki/Special:BookSources/0-444-89264-8http://online.liebertpub.com/doi/abs/10.1089/106652700750050961http://books.google.com/books?id=VkawQgAACAAJhttp://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://en.wikipedia.org/wiki/Digital_object_identifierhttp://ftp.cs.ucla.edu/pub/stat_ser/r116.pdfhttp://en.wikipedia.org/wiki/Special:BookSources/978-0-13-012534-7http://link.springer.com/article/10.1023%2FA%3A1007465528199http://en.wikipedia.org/wiki/PubMed_Identifierhttp://en.wikipedia.org/wiki/International_Standard_Book_Number
  • 5/21/2015 BayesiannetworkWikipedia,thefreeencyclopedia

    http://en.wikipedia.org/wiki/Bayesian_network 12/15

    22. J.Uebersax(2004).GeneticCounselingandCancerRiskModeling:AnApplicationofBayesNets(http://www.johnuebersax.com/stat/bayes_net_breast_cancer.doc).Marbella,Spain:RavenpackInternational.

    23. JiangX,CooperGF.(JulyAugust2010)."ABayesianspatiotemporalmethodfordiseaseoutbreakdetection"(http://jamia.bmj.com/cgi/pmidlookup?view=long&pmid=20595315).JAmMedInformAssoc17(4):46271.doi:10.1136/jamia.2009.000356(https://dx.doi.org/10.1136%2Fjamia.2009.000356).PMC2995651(https://www.ncbi.nlm.nih.gov/pmc/articles/PMC2995651).PMID20595315(https://www.ncbi.nlm.nih.gov/pubmed/20595315).

    24. LuisM.deCampos,JuanM.FernndezLunaandJuanF.Huete(2004)."Bayesiannetworksandinformationretrieval:anintroductiontothespecialissue".InformationProcessing&Management(Elsevier)40(5):727733.doi:10.1016/j.ipm.2004.03.001(https://dx.doi.org/10.1016%2Fj.ipm.2004.03.001).ISBN0471141828.

    25. ChristosL.KoumenidesandNigelR.Shadbolt.2012.CombininglinkandcontentbasedinformationinaBayesianinferencemodelforentitysearch.(http://eprints.soton.ac.uk/342220)InProceedingsofthe1stJointInternationalWorkshoponEntityOrientedandSemanticSearch(JIWES'12).ACM,NewYork,NY,USA,,Article3,6pages.doi:10.1145/2379307.2379310(https://dx.doi.org/10.1145%2F2379307.2379310)

    26. F.J.Dez,J.Mira,E.IturraldeandS.Zubillaga(1997)."DIAVAL,aBayesianexpertsystemforechocardiography"(http://www.cisiad.uned.es/papers/diaval.php).ArtificialIntelligenceinMedicine(Elsevier)10(1):5973.doi:10.1016/s09333657(97)003849(https://dx.doi.org/10.1016%2Fs09333657%2897%29003849).PMID9177816(https://www.ncbi.nlm.nih.gov/pubmed/9177816).

    27. Constantinou,AnthonyFenton,N.Neil,M.(2012)."pifootball:ABayesiannetworkmodelforforecastingAssociationFootballmatchoutcomes"(http://dx.doi.org/10.1016/j.knosys.2012.07.008).KnowledgeBasedSystems36:322339.doi:10.1016/j.knosys.2012.07.008(https://dx.doi.org/10.1016%2Fj.knosys.2012.07.008).Retrieved25March2014.

    28. Constantinou,AnthonyFenton,N.Neil,M.(2013)."ProfitingfromaninefficientAssociationFootballgamblingmarket:Prediction,RiskandUncertaintyusingBayesiannetworks."(http://dx.doi.org/10.1016/j.knosys.2013.05.008).KnowledgeBasedSystems50:6086.doi:10.1016/j.knosys.2013.05.008(https://dx.doi.org/10.1016%2Fj.knosys.2013.05.008).Retrieved25March2014.

    29. G.A.Davis(2003)."Bayesianreconstructionoftrafficaccidents".Law,ProbabilityandRisk2(2):6989.doi:10.1093/lpr/2.2.69(https://dx.doi.org/10.1093%2Flpr%2F2.2.69).

    30. J.B.KadaneandD.A.Schum(1996).AProbabilisticAnalysisoftheSaccoandVanzettiEvidence.NewYork:Wiley.ISBN0471141828.

    31. O.Pourret,P.NaimandB.Marcot(2008).BayesianNetworks:APracticalGuidetoApplications(http://www.wiley.com/go/pourret).Chichester,UK:Wiley.ISBN9780470060308.

    32. Karvanen,Juha(2014)."Studydesignincausalmodels".ScandinavianJournalofStatistics.doi:10.1111/sjos.12110(https://dx.doi.org/10.1111%2Fsjos.12110).

    33. Cardenas,ICAlJibouri,SHSHalman,JIMvanTol,FA(2014)."ModelingRiskRelatedKnowledgeinTunnelingProjects"(http://onlinelibrary.wiley.com/doi/10.1111/risa.12094/abstract?deniedAccessCustomisedMessage=&userIsAuthenticated=false).RiskAnalysis34(2):323339.doi:10.1111/risa.12094(https://dx.doi.org/10.1111%2Frisa.12094).

    34. Cardenas,ICAlJibouri,SHSHalman,JIMvandeLinde,WKaalberg,F(2014)."UsingPriorRiskRelatedKnowledgetoSupportRiskManagementDecisions:LessonsLearntfromaTunnelingProject"(http://onlinelibrary.wiley.com/doi/10.1111/risa.12213/abstract).RiskAnalysis34(8).doi:10.1111/risa.12213(https://dx.doi.org/10.1111%2Frisa.12213).

    35. Cardenas,IC(2015)."ModelingtheInfluenceofUnknownFactorsinRiskAnalysisUsingBayesianNetworks"(http://www.researchgate.net/publication/274456154_Modeling_the_Influence_of_Unknown_Factors_in_Risk_Analysis_using_Bayesian_Networks).Underreviewbyarefereedjournal.

    36. Neapolitan,Richard(2009).ProbabilisticMethodsforBioinformatics(https://www.elsevier.com/books/probabilisticmethodsforbioinformatics/neapolitan/9780123704764).Burlington,MA:MorganKaufmann.p.406.ISBN9780123704764.

    37. Neapolitan,Richard,andXiaJiang(2007).ProbabilisticMethodsforFinancialandMarketingInformatics(http://store.elsevier.com/ProbabilisticMethodsforFinancialandMarketingInformatics/RichardE_Neapolitan/isbn9780123704771/).Burlingon,MA:MorganKaufmann.p.432.ISBN0123704774.

    http://en.wikipedia.org/wiki/Special:BookSources/0-471-14182-8http://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://dx.doi.org/10.1016/j.knosys.2012.07.008http://dx.doi.org/10.1136%2Fjamia.2009.000356http://dx.doi.org/10.1016%2Fj.ipm.2004.03.001http://dx.doi.org/10.1016%2Fj.knosys.2013.05.008http://dx.doi.org/10.1016%2Fs0933-3657%2897%2900384-9http://en.wikipedia.org/wiki/Digital_object_identifierhttp://www.researchgate.net/publication/274456154_Modeling_the_Influence_of_Unknown_Factors_in_Risk_Analysis_using_Bayesian_Networkshttp://en.wikipedia.org/wiki/Special:BookSources/0123704774http://en.wikipedia.org/wiki/PubMed_Identifierhttp://en.wikipedia.org/wiki/Special:BookSources/978-0-470-06030-8http://en.wikipedia.org/wiki/PubMed_Centralhttp://en.wikipedia.org/wiki/Digital_object_identifierhttp://dx.doi.org/10.1016/j.knosys.2013.05.008http://en.wikipedia.org/wiki/Digital_object_identifierhttp://en.wikipedia.org/wiki/PubMed_Identifierhttp://dx.doi.org/10.1111%2Frisa.12213http://dx.doi.org/10.1016%2Fj.knosys.2012.07.008http://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://www.wiley.com/go/pourrethttp://dx.doi.org/10.1093%2Flpr%2F2.2.69http://eprints.soton.ac.uk/342220http://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://en.wikipedia.org/wiki/Special:BookSources/9780123704764http://jamia.bmj.com/cgi/pmidlookup?view=long&pmid=20595315http://www.ncbi.nlm.nih.gov/pubmed/9177816http://en.wikipedia.org/wiki/Digital_object_identifierhttp://store.elsevier.com/Probabilistic-Methods-for-Financial-and-Marketing-Informatics/Richard-E_-Neapolitan/isbn-9780123704771/http://en.wikipedia.org/wiki/Digital_object_identifierhttp://onlinelibrary.wiley.com/doi/10.1111/risa.12094/abstract?deniedAccessCustomisedMessage=&userIsAuthenticated=falsehttp://www.john-uebersax.com/stat/bayes_net_breast_cancer.dochttp://dx.doi.org/10.1145%2F2379307.2379310http://dx.doi.org/10.1111%2Frisa.12094http://en.wikipedia.org/wiki/Digital_object_identifierhttp://dx.doi.org/10.1111%2Fsjos.12110http://en.wikipedia.org/wiki/Special:BookSources/0-471-14182-8http://www.ncbi.nlm.nih.gov/pmc/articles/PMC2995651http://en.wikipedia.org/wiki/Digital_object_identifierhttp://en.wikipedia.org/wiki/Digital_object_identifierhttp://en.wikipedia.org/wiki/International_Standard_Book_Numberhttps://www.elsevier.com/books/probabilistic-methods-for-bioinformatics/neapolitan/978-0-12-370476-4http://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://www.cisiad.uned.es/papers/diaval.phphttp://en.wikipedia.org/wiki/Digital_object_identifierhttp://www.ncbi.nlm.nih.gov/pubmed/20595315http://en.wikipedia.org/wiki/Digital_object_identifierhttp://onlinelibrary.wiley.com/doi/10.1111/risa.12213/abstract
  • 5/21/2015 BayesiannetworkWikipedia,thefreeencyclopedia

    http://en.wikipedia.org/wiki/Bayesian_network 13/15

    References

    38. Pearl,J.(1985).BayesianNetworks:AModelofSelfActivatedMemoryforEvidentialReasoning(http://ftp.cs.ucla.edu/techreport/198_reports/850017.pdf)(UCLATECHNICALREPORTCSD850017).Proceedingsofthe7thConferenceoftheCognitiveScienceSociety,UniversityofCalifornia,Irvine,CA.pp.329334.Retrieved20090501.

    39. Bayes,T.Price,Mr.(1763)."AnEssaytowardssolvingaProblemintheDoctrineofChances".PhilosophicalTransactionsoftheRoyalSociety53:370418.doi:10.1098/rstl.1763.0053(https://dx.doi.org/10.1098%2Frstl.1763.0053).

    40. Pearl,J.ProbabilisticReasoninginIntelligentSystems(http://books.google.com/books?id=AvNID7LyMusC).SanFranciscoCA:MorganKaufmann.p.1988.ISBN1558604790.

    41. Neapolitan,RichardE.(1989).Probabilisticreasoninginexpertsystems:theoryandalgorithms(http://www.amazon.com/ProbabilisticReasoningExpertSystemsAlgorithms/dp/1477452540/ref=sr_1_3?s=books&ie=UTF8&qid=1389578837&sr=13&keywords=probabilistic+reasoning+in+expert+systems).Wiley.ISBN9780471618409.

    42. Wright,S.(1921)."CorrelationandCausation"(http://www.ssc.wisc.edu/soc/class/soc952/Wright/Wright_Correlation%20and%20Causation.pdf)(PDF).JournalofAgriculturalResearch20(7):557585.

    BenGal,Irad(2007)."BayesianNetworks".InRuggeri,FabrizioKennett,RonS.Faltin,FrederickW.EncyclopediaofStatisticsinQualityandReliability(http://www.eng.tau.ac.il/~bengal/BN.pdf)(PDF).EncyclopediaofStatisticsinQualityandReliability.JohnWiley&Sons.doi:10.1002/9780470061572.eqr089(https://dx.doi.org/10.1002%2F9780470061572.eqr089).ISBN9780470018613.BertschMcGrayne,Sharon.TheTheoryThatWouldnotDie.Yale.Borgelt,ChristianKruse,Rudolf(March2002).GraphicalModels:MethodsforDataAnalysisandMining(http://fuzzy.cs.unimagdeburg.de/books/gm/).Chichester,UK:Wiley.ISBN0470843373.Borsuk,MarkEdward(2008)."Ecologicalinformatics:Bayesiannetworks".InJrgensen,SvenErik,Fath,Brian.EncyclopediaofEcology.Elsevier.ISBN9780444520333.Cardenas,I.etal.(April2015)."ModelingtheInfluenceofUnknownFactorsinRiskAnalysisusingBayesianNetworks"(http://www.researchgate.net/publication/274456154_Modeling_the_Influence_of_Unknown_Factors_in_Risk_Analysis_using_Bayesian_Networks)(PDF).Underreviewbyarefereedjournal.Castillo,EnriqueGutirrez,JosManuelHadi,AliS.(1997)."LearningBayesianNetworks".ExpertSystemsandProbabilisticNetworkModels.Monographsincomputerscience.NewYork:SpringerVerlag.pp.481528.ISBN0387948589.Comley,JoshuaW.Dowe,DavidL.(http://www.csse.monash.edu.au/~dld)(October2003)."MinimumMessageLengthandGeneralizedBayesianNetswithAsymmetricLanguages"(http://www.csse.monash.edu.au/~dld/David.Dowe.publications.html#ComleyDowe2005).WrittenatVictoria,Australia.InGrnwald,PeterD.Myung,InJaePitt,MarkA.AdvancesinMinimumDescriptionLength:TheoryandApplications.Neuralinformationprocessingseries.Cambridge,Massachusetts:BradfordBooks(MITPress)(publishedApril2005).pp.265294.ISBN0262072629.(ThispaperputsdecisiontreesininternalnodesofBayesnetworksusingMinimumMessageLength(http://www.csse.monash.edu.au/~dld/MML.html)(MML).AnearlierversionisComleyandDowe(2003)(http://www.csse.monash.edu.au/~dld/David.Dowe.publications.html#ComleyDowe2003),.pdf(http://www.csse.monash.edu.au/~dld/Publications/2003/Comley+Dowe03_HICS2003_GeneralBayesianNetworksAsymmetricLanguages.pdf).)Darwiche,Adnan(2009).ModelingandReasoningwithBayesianNetworks(http://www.cambridge.org/9780521884389).CambridgeUniversityPress.ISBN9780521884389.

    http://www.cambridge.org/9780521884389http://en.wikipedia.org/wiki/Special:BookSources/978-0-444-52033-3http://en.wikipedia.org/wiki/Springer_Science%2BBusiness_Mediahttp://www.csse.monash.edu.au/~dld/David.Dowe.publications.html#ComleyDowe2005http://en.wikipedia.org/wiki/Digital_object_identifierhttp://www.csse.monash.edu.au/~dld/MML.htmlhttp://en.wikipedia.org/wiki/Philosophical_Transactions_of_the_Royal_Societyhttp://en.wikipedia.org/wiki/Special:BookSources/978-0521884389http://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://en.wikipedia.org/wiki/Decision_tree_learninghttp://ftp.cs.ucla.edu/tech-report/198_-reports/850017.pdfhttp://en.wikipedia.org/wiki/Special:BookSources/978-0-470-01861-3http://en.wikipedia.org/wiki/Yalehttp://www.amazon.com/Probabilistic-Reasoning-Expert-Systems-Algorithms/dp/1477452540/ref=sr_1_3?s=books&ie=UTF8&qid=1389578837&sr=1-3&keywords=probabilistic+reasoning+in+expert+systemshttp://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://books.google.com/books?id=AvNID7LyMusChttp://en.wikipedia.org/wiki/Special:BookSources/1558604790http://en.wikipedia.org/wiki/Special:BookSources/0-470-84337-3http://en.wikipedia.org/wiki/Judea_Pearlhttp://en.wikipedia.org/wiki/MIT_Presshttp://www.csse.monash.edu.au/~dld/David.Dowe.publications.html#ComleyDowe2003http://www.researchgate.net/publication/274456154_Modeling_the_Influence_of_Unknown_Factors_in_Risk_Analysis_using_Bayesian_Networkshttp://en.wikipedia.org/wiki/Special:BookSources/978-0-471-61840-9http://en.wikipedia.org/wiki/Sven_Erik_J%C3%B8rgensenhttp://en.wikipedia.org/wiki/John_Wiley_%26_Sonshttp://dx.doi.org/10.1002%2F9780470061572.eqr089http://www.eng.tau.ac.il/~bengal/BN.pdfhttp://www.csse.monash.edu.au/~dld/Publications/2003/Comley+Dowe03_HICS2003_GeneralBayesianNetworksAsymmetricLanguages.pdfhttp://en.wikipedia.org/wiki/Victoria_(Australia)http://www.csse.monash.edu.au/~dldhttp://en.wikipedia.org/wiki/Chichesterhttp://en.wikipedia.org/wiki/Minimum_message_lengthhttp://fuzzy.cs.uni-magdeburg.de/books/gm/http://en.wikipedia.org/wiki/John_Wiley_%26_Sonshttp://en.wikipedia.org/wiki/Thomas_Bayeshttp://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://www.ssc.wisc.edu/soc/class/soc952/Wright/Wright_Correlation%20and%20Causation.pdfhttp://en.wikipedia.org/wiki/Digital_object_identifierhttp://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://en.wikipedia.org/wiki/An_Essay_towards_solving_a_Problem_in_the_Doctrine_of_Chanceshttp://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://en.wikipedia.org/wiki/Sewall_Wrighthttp://en.wikipedia.org/wiki/Cambridge_University_Presshttp://en.wikipedia.org/wiki/Special:BookSources/0-262-07262-9http://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://en.wikipedia.org/wiki/Special:BookSources/0-387-94858-9http://dx.doi.org/10.1098%2Frstl.1763.0053http://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://en.wikipedia.org/wiki/Cambridge,_Massachusetts
  • 5/21/2015 BayesiannetworkWikipedia,thefreeencyclopedia

    http://en.wikipedia.org/wiki/Bayesian_network 14/15

    Dowe,DavidL.(2010).MML,hybridBayesiannetworkgraphicalmodels,statisticalconsistency,invarianceanduniqueness(http://www.csse.monash.edu.au/~dld/Publications/2010/Dowe2010_MML_HandbookPhilSci_Vol7_HandbookPhilStat_MML+hybridBayesianNetworkGraphicalModels+StatisticalConsistency+InvarianceAndUniqueness_pp901982.pdf),inHandbookofPhilosophyofScience(Volume7:HandbookofPhilosophyofStatistics),Elsevier,ISBN9780444518620(http://japan.elsevier.com/products/books/HPS.pdf),pp901982(http://www.csse.monash.edu.au/~dld/Publications/2010/Dowe2010_MML_HandbookPhilSci_Vol7_HandbookPhilStat_MML+hybridBayesianNetworkGraphicalModels+StatisticalConsistency+InvarianceAndUniqueness_pp901982.pdf).Fenton,NormanNeil,MartinE.(November2007).ManagingRiskintheModernWorld:ApplicationsofBayesianNetworks(http://www.agenarisk.com/resources/apps_bayesian_networks.pdf)AKnowledgeTransferReportfromtheLondonMathematicalSocietyandtheKnowledgeTransferNetworkforIndustrialMathematics.London(England):LondonMathematicalSociety.Fenton,NormanNeil,MartinE.(July23,2004)."CombiningevidenceinriskanalysisusingBayesianNetworks"(https://www.dcs.qmul.ac.uk/~norman/papers/Combining%20evidence%20in%20risk%20analysis%20using%20BNs.pdf)(PDF).SafetyCriticalSystemsClubNewsletter13(4)(NewcastleuponTyne,England).pp.813.AndrewGelmanJohnBCarlinHalSSternDonaldBRubin(2003)."PartII:FundamentalsofBayesianDataAnalysis:Ch.5Hierarchicalmodels"(http://books.google.com/books?id=TNYhnkXQSjAC&pg=PA120).BayesianDataAnalysis(http://books.google.com.au/books?id=TNYhnkXQSjAC).CRCPress.pp.120.ISBN9781584883883.Heckerman,David(March1,1995)."TutorialonLearningwithBayesianNetworks"(http://research.microsoft.com/research/pubs/view.aspx?msr_tr_id=MSRTR9506).InJordan,MichaelIrwin.LearninginGraphicalModels.AdaptiveComputationandMachineLearning.Cambridge,Massachusetts:MITPress(published1998).pp.301354.ISBN0262600323..

    AlsoappearsasHeckerman,David(March1997)."BayesianNetworksforDataMining".DataMiningandKnowledgeDiscovery(Netherlands:SpringerNetherlands)1(1):79119.doi:10.1023/A:1009730122752(https://dx.doi.org/10.1023%2FA%3A1009730122752).ISSN13845810(https://www.worldcat.org/issn/13845810).AnearlierversionappearsasTechnicalReportMSRTR9506(http://research.microsoft.com/research/pubs/view.aspx?msr_tr_id=MSRTR9506),MicrosoftResearch,March1,1995.ThepaperisaboutbothparameterandstructurelearninginBayesiannetworks.

    Jensen,FinnVNielsen,ThomasD.(June6,2007).BayesianNetworksandDecisionGraphs.InformationScienceandStatisticsseries(2nded.).NewYork:SpringerVerlag.ISBN9780387682815.Karimi,KamranHamilton,HowardJ.(2000)."Findingtemporalrelations:Causalbayesiannetworksvs.C4.5"(http://www.kamrankarimi.com/pubs/khISMIS2000.pdf)(PDF).TwelfthInternationalSymposiumonMethodologiesforIntelligentSystems.Korb,KevinB.Nicholson,AnnE.(December2010).BayesianArtificialIntelligence.CRCComputerScience&DataAnalysis(2nded.).Chapman&Hall(CRCPress).doi:10.1007/s1004400402145(https://dx.doi.org/10.1007%2Fs1004400402145).ISBN1584883871.Lunn,D.Thomas,ABest,Netal.(2009)."TheBUGSproject:Evolution,critiqueandfuturedirections".StatisticsinMedicine28(25):30493067.doi:10.1002/sim.3680(https://dx.doi.org/10.1002%2Fsim.3680).PMID19630097(https://www.ncbi.nlm.nih.gov/pubmed/19630097).|first2=missing|last2=inAuthorslist(help)Neil,MartinFenton,NormanE.Tailor,Manesh(August2005).Greenberg,MichaelR.,ed."UsingBayesianNetworkstoModelExpectedandUnexpectedOperationalLosses"(http://www.dcs.qmul.ac.uk/~norman/papers/oprisk.pdf)(PDF).RiskAnalysis:anInternationalJournal(JohnWiley&Sons)25(4):963972.doi:10.1111/j.15396924.2005.00641.x(https://dx.doi.org/10.1111%2Fj.15396924.2005.00641.x).PMID16268944(https://www.ncbi.nlm.nih.gov/pubmed/16268944).Pearl,Judea(September1986)."Fusion,propagation,andstructuringinbeliefnetworks".ArtificialIntelligence(Elsevier)29(3):241288.doi:10.1016/00043702(86)90072X(https://dx.doi.org/10.1016%2F00043702%2886%2990072X).ISSN00043702(https://www.worldcat.org/issn/00043702).

    http://en.wikipedia.org/wiki/International_Standard_Serial_Numberhttp://en.wikipedia.org/wiki/John_Wiley_%26_Sonshttp://research.microsoft.com/research/pubs/view.aspx?msr_tr_id=MSR-TR-95-06http://en.wikipedia.org/wiki/Help:CS1_errors#first_missing_lasthttps://www.dcs.qmul.ac.uk/~norman/papers/Combining%20evidence%20in%20risk%20analysis%20using%20BNs.pdfhttp://en.wikipedia.org/wiki/Digital_object_identifierhttp://en.wikipedia.org/wiki/CRC_Presshttp://en.wikipedia.org/wiki/Special:BookSources/0-262-60032-3http://www.ncbi.nlm.nih.gov/pubmed/16268944http://en.wikipedia.org/wiki/Digital_object_identifierhttp://www.worldcat.org/issn/0004-3702http://en.wikipedia.org/wiki/Springer_Science%2BBusiness_Mediahttp://en.wikipedia.org/wiki/Chapman_%26_Hallhttp://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://dx.doi.org/10.1111%2Fj.1539-6924.2005.00641.xhttp://books.google.com/books?id=TNYhnkXQSjAC&pg=PA120http://en.wikipedia.org/wiki/Cambridge,_Massachusettshttp://en.wikipedia.org/wiki/Elsevierhttp://www.csse.monash.edu.au/~dld/Publications/2010/Dowe2010_MML_HandbookPhilSci_Vol7_HandbookPhilStat_MML+hybridBayesianNetworkGraphicalModels+StatisticalConsistency+InvarianceAndUniqueness_pp901-982.pdfhttp://en.wikipedia.org/wiki/MIT_Presshttp://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://en.wikipedia.org/wiki/Artificial_Intelligence_(journal)http://en.wikipedia.org/wiki/Digital_object_identifierhttp://dx.doi.org/10.1023%2FA%3A1009730122752http://www.ncbi.nlm.nih.gov/pubmed/19630097http://en.wikipedia.org/wiki/International_Standard_Serial_Numberhttp://en.wikipedia.org/wiki/PubMed_Identifierhttp://en.wikipedia.org/wiki/Special:BookSources/1-58488-387-1http://en.wikipedia.org/wiki/London_Mathematical_Societyhttp://japan.elsevier.com/products/books/HPS.pdfhttp://en.wikipedia.org/wiki/Data_Mining_and_Knowledge_Discoveryhttp://research.microsoft.com/research/pubs/view.aspx?msr_tr_id=MSR-TR-95-06http://en.wikipedia.org/wiki/Digital_object_identifierhttp://www.dcs.qmul.ac.uk/~norman/papers/oprisk.pdfhttp://www.agenarisk.com/resources/apps_bayesian_networks.pdfhttp://en.wikipedia.org/wiki/Digital_object_identifierhttp://www.csse.monash.edu.au/~dld/Publications/2010/Dowe2010_MML_HandbookPhilSci_Vol7_HandbookPhilStat_MML+hybridBayesianNetworkGraphicalModels+StatisticalConsistency+InvarianceAndUniqueness_pp901-982.pdfhttp://en.wikipedia.org/wiki/Judea_Pearlhttp://www.worldcat.org/issn/1384-5810http://www.kamran-karimi.com/pubs/khISMIS2000.pdfhttp://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://en.wikipedia.org/wiki/Netherlandshttp://dx.doi.org/10.1007%2Fs10044-004-0214-5http://en.wikipedia.org/wiki/PubMed_Identifierhttp://en.wikipedia.org/wiki/Society_for_Risk_Analysishttp://dx.doi.org/10.1002%2Fsim.3680http://books.google.com.au/books?id=TNYhnkXQSjAChttp://en.wikipedia.org/wiki/Special:BookSources/978-0-387-68281-5http://en.wikipedia.org/wiki/Springer_Science%2BBusiness_Mediahttp://en.wikipedia.org/wiki/Londonhttp://dx.doi.org/10.1016%2F0004-3702%2886%2990072-Xhttp://en.wikipedia.org/wiki/New_Yorkhttp://en.wikipedia.org/wiki/Newcastle_upon_Tynehttp://en.wikipedia.org/wiki/Special:BookSources/978-1-58488-388-3
  • 5/21/2015 BayesiannetworkWikipedia,thefreeencyclopedia

    http://en.wikipedia.org/wiki/Bayesian_network 15/15

    Furtherreading

    ComputationalIntelligence:AMethodologicalIntroductionbyKruse,Borgelt,Klawonn,Moewes,Steinbrecher,Held,2013,Springer,ISBN9781447150121GraphicalModelsRepresentationsforLearning,ReasoningandDataMining,2ndEdition,byBorgelt,Steinbrecher,Kruse,2009,J.Wiley&Sons,ISBN9780470749562

    Externallinks

    AtutorialonlearningwithBayesianNetworks(http://research.microsoft.com/research/pubs/view.aspx?msr_tr_id=MSRTR9506)AnIntroductiontoBayesianNetworksandtheirContemporaryApplications(http://www.niedermayer.ca/papers/bayesian/bayes.html)OnlineTutorialonBayesiannetsandprobability(http://www.dcs.qmw.ac.uk/%7Enorman/BBNs/BBNs.htm)WebApptocreateBayesiannetsandrunitwithaMonteCarlomethod(http://princesofserendib.com/)ContinuousTimeBayesianNetworks(http://robotics.stanford.edu/~nodelman/papers/ctbn.pdf)BayesianNetworks:ExplanationandAnalogy(http://wiki.syncleus.com/index.php/DANN:Bayesian_Network)AlivetutorialonlearningBayesiannetworks(http://videolectures.net/kdd07_neapolitan_lbn/)AhierarchicalBayesModelforhandlingsampleheterogeneityinclassificationproblems(http://www.biomedcentral.com/14712105/7/514/abstract),providesaclassificationmodeltakingintoconsiderationtheuncertaintyassociatedwithmeasuringreplicatesamples.HierarchicalNaiveBayesModelforhandlingsampleuncertainty(http://www.labmedinfo.org/download/lmi339.pdf),showshowtoperformclassificationandlearningwithcontinuousanddiscretevariableswithreplicatedmeasurements.

    Retrievedfrom"http://en.wikipedia.org/w/index.php?title=Bayesian_network&oldid=660234761"

    Categories: Bayesiannetworks Networks Statisticalmodels Graphicalmodels

    Thispagewaslastmodifiedon1May2015,at12:22.TextisavailableundertheCreativeCommonsAttributionShareAlikeLicenseadditionaltermsmayapply.Byusingthissite,youagreetotheTermsofUseandPrivacyPolicy.WikipediaisaregisteredtrademarkoftheWikimediaFoundation,Inc.,anonprofitorganization.

    Pearl,Judea(1988).ProbabilisticReasoninginIntelligentSystems:NetworksofPlausibleInference.RepresentationandReasoningSeries(2ndprintinged.).SanFrancisco,California:MorganKaufmann.ISBN0934613737.Pearl,JudeaRussell,Stuart(November2002)."BayesianNetworks".InArbib,MichaelA.HandbookofBrainTheoryandNeuralNetworks.Cambridge,Massachusetts:BradfordBooks(MITPress).pp.157160.ISBN0262011972.Russell,StuartJ.Norvig,Peter(2003),ArtificialIntelligence:AModernApproach(http://aima.cs.berkeley.edu/)(2nded.),UpperSaddleRiver,NewJersey:PrenticeHall,ISBN0137903952.Zhang,NevinLianwen(http://www.cs.ust.hk/faculty/lzhang/bio.html)Poole,David(http://www.cs.ubc.ca/spider/poole/)(May1994)."AsimpleapproachtoBayesiannetworkcomputations".ProceedingsoftheTenthBiennialCanadianArtificialIntelligenceConference(AI94).(Banff,Alberta):171178.Thispaperpresentsvariableeliminationforbeliefnetworks.

    http://www.cs.ust.hk/faculty/lzhang/bio.htmlhttp://en.wikipedia.org/wiki/Judea_Pearlhttp://wikimediafoundation.org/wiki/Privacy_policyhttp://wiki.syncleus.com/index.php/DANN:Bayesian_Networkhttp://en.wikipedia.org/wiki/Stuart_J._Russellhttp://en.wikipedia.org/wiki/Category:Bayesian_networkshttp://en.wikipedia.org/wiki/Stuart_J._Russellhttp://en.wikipedia.org/wiki/Peter_Norvighttp://en.wikipedia.org/wiki/Category:Networkshttp://en.wikipedia.org/wiki/Category:Graphical_modelshttp://en.wikipedia.org/wiki/Special:BookSources/0-13-790395-2http://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://en.wikipedia.org/wiki/Banff,_Albertahttp://en.wikipedia.org/wiki/Special:BookSources/0-262-01197-2http://en.wikipedia.org/wiki/Help:Categoryhttp://en.wikipedia.org/wiki/Judea_Pearlhttp://en.wikipedia.org/wiki/Wikipedia:Text_of_Creative_Commons_Attribution-ShareAlike_3.0_Unported_Licensehttp://en.wikipedia.org/wiki/Cambridge,_Massachusettshttp://en.wikipedia.org/wiki/Category:Statistical_modelshttp://wikimediafoundation.org/wiki/Terms_of_Usehttp://www.dcs.qmw.ac.uk/~norman/BBNs/BBNs.htmhttp://www.biomedcentral.com/1471-2105/7/514/abstracthttp://aima.cs.berkeley.edu/http://princesofserendib.com/http://www.cs.ubc.ca/spider/poole/http://en.wikipedia.org/wiki/San_Francisco,_Californiahttp://videolectures.net/kdd07_neapolitan_lbn/http://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://en.wikipedia.org/wiki/Special:BookSources/9780470749562http://www.labmedinfo.org/download/lmi339.pdfhttp://en.wikipedia.org/wiki/Special:BookSources/0-934613-73-7http://research.microsoft.com/research/pubs/view.aspx?msr_tr_id=MSR-TR-95-06http://en.wikipedia.org/w/index.php?title=Canadian_Conference_on_Artificial_Intelligence&action=edit&redlink=1http://en.wikipedia.org/wiki/Michael_A._Arbibhttp://en.wikipedia.org/w/index.php?title=Bayesian_network&oldid=660234761http://www.wikimediafoundation.org/http://en.wikipedia.org/wiki/Morgan_Kaufmannhttp://en.wikipedia.org/wiki/Special:BookSources/9781447150121http://www.niedermayer.ca/papers/bayesian/bayes.htmlhttp://en.wikipedia.org/wiki/MIT_Presshttp://en.wikipedia.org/wiki/International_Standard_Book_Numberhttp://robotics.stanford.edu/~nodelman/papers/ctbn.pdf