简介：Inthispaper,themethodsbasedonlikelihoodratioareproposedfeeuncertaintyreasoning.Theelementaryconceptsandformulasarederivedsysttmatically,especiellyforthesituationswithdependentinformationandforthedifferenrordersofinformalion.

简介：

简介：ThepropertyofNPcompletenessoftopologicspatialreasoningproblemhasbeenproved.Accordingtothesimilarityofuncertaintywithtopologicspatialreasoning,theproblemofdirectionalspatialreasoningshouldbealsoanNPcompleteproblem.TheproofforthepropertyofNPcompletenessindirectionalspatialreasoningproblemisbasedontwoimportanttransformations.Afterthesetransformations,aspatialconfigurationhasbeenconstructedbasedondirectionalconstraints,andthepropertyofNPcompletenessindirectionalspatialreasoninghasbeenprovedwiththehelpoftheconsistencyoftheconstraintsintheconfiguration.

简介：

简介：

简介：AsanimportantvariantofReiter'sdefaultlogic.Poole(1988)developedanonmonotonicreasoningframeworkintheclassicalfirst-orderlanguage,BrewkaandNebelextendedPoole'sapproachinordertoenablearepresentationofprioritiesbetweendefaults.Inthispaperageneralframeworkfordefaultreasoningispresented,whichcanbeviewedasageneralizationofthethreeapproachesabove.Itisprovedthatthesyntax-independentdefaultreasoninginthisframeworkisidenticaltothegeneralbeliefrevisionoperationintroducedbyZhangetal.(1997).Thisesultprovidesasolutiontotheproblemwhetherthereisacorrespondencebetweenbeliefrevisionanddefaultlogicfortheinfinitecase.Asaby-product,ananswertothethequestion,raisedbyMankinsonandGaerdenfors(1991),isalsogivenaboutwhetherthereisacounterpartcontracitoninnonmonotoniclogic.

简介：ＡＦｕｚｚｙＲｅａｓｏｎｉｎｇＳｙｓｔｅｍａｎｄＩｔｓＨｅｕｒｉｓｔｉｃＩｎｆｅｒｅｎｃｅＡｌｇｏｒｉｔｈｍＺｕｏＸｉａｏｄｅ＆ＬｉａｎｇＹｕｎＤｅｐｔ．ｏｆＢｕｓｉｎｅｓＡｄｍｉｎｉｓｔｒａｔｉｏｎ，ＪｉｎａｎＵｎｉｖｅｒｓｉｔｙ，Ｇｕａｎｇｚｈｏ...

简介：基于盒子的推理是以前的答案在被存储为未来使用的一种AI技术。人们习惯于根据在他们的记忆被存储并且以前被他们使用了的那些线路指导自己。只基于人到熟悉的线路的偏爱，它通过认知活动的学习被获得。我们建议基于推理到路径计划的盒子使用聪明的方法。一个盒子底存储所有答案到所有最短的路径是不可能的；因此，他们的部分应该被存储。不管多么，线路哪个并且哪个应该被存储不应该是？我们怎么并且怎么改编已经被存储了的盒子我们基于他们获得最短的线路？所有这些问题需要被由于基于盒子的推理技术，集成网络的知识解释。这篇论文建议在另一个点的基于盒子的推理。根据问题空间的总分析发现一些不能替代的连接的这个工具，它被称为must_be_passed在来源和目的地之间连接。仅仅从格子的那些最好的出口/入口节点计算最短的路径盒子到不能替代的连接，然后增加他们进存储为未来使用的盒子底。这个方法基于基于盒子的推理技术并且完全考虑问题空间的性质。除了在线路网络的自然格子的知识的使用，这个方法是比计算效率上的存在算法更有效的。

简介：Inmanyapplicationsanddomains,temporalconstraintsbetweenactions,andtheirprobabilitiesplayanimportantrole.Weproposethefirstapproachintheliteraturecopingwithprobabilisticquantitativeconstraints.Toachievesuchachallenginggoal,weextendthewidelyusedsimpletemporalproblem(STP)frameworktoconsiderprobabilities.Specifically,weproposei)aformalrepresentationofprobabilisticquantitativeconstraints,ii)analgorithm,basedontheoperationsofintersectionandcomposition,forthepropagationofsuchtemporalconstraints,andiii)facilitiestosupportqueryansweringonasetofsuchconstraints.Asaresult,weprovideuserswiththefirsthomogeneousmethodsupportingthetreatment(representing,reasoning,andquerying)ofprobabilisticquantitativeconstraints,asrequiredbymanyapplicationsanddomains.

简介：Tousereasoningknowledgeaccuratelyandefficiently,manyreasoningmethodshavebeenproposed.However,thedifferencesinformamongthemethodsmayobstructthesystematicalanalysisandharmoniousintegrationofthem.Inthispaper,acommonreasoningmodelJUM(JudgementModel)isintroduced.AccordingtoJUM,acommonknowledgerepresentationformisabstractedfromdifferentreasoningmethodsanditslimitationisreduced.WealsoproposeanalgorithmfortransformingonetypeofJUMsintoanother.Insomecases,thealgorithmcanbeusedtoresolvethekeyproblemofintegratingdifferenttypesofJUMinonesystem.Itispossiblethatanewarchitectureofknowledge-basedsystemcanberealizedunderJUM.

简介：

简介：

简介：AfteranalyzingtheweldingprocedureknowledgeinChinesenationalstandardsforweldingprocedurequalificationofsteelpressurevesselfromthepointofestablishingexpertsystem,itcanbedividedintofivetypesofknowledge,i.e.practice,definition,regularity,processanddescriptionknowledge.Theknowledgeexpressionmethodsareestablishedaccordingtothedifferenttypeofweldingprocedureknowledge.Thereasoningprocessbasedonruleisadopted.Andthereasoningengineisembeddedamongobjectsintegratedwiththeknowledgebase.

简介：

简介：Heuristicisanimportantmethodtosolvetheshortestpathproblems.Itindicatesliterallylearningbyexperience,ormoregenerallyintheartificialintelligenceliterature,aheuristicisa“ruleofthumb”andassuchistheapproachusedbyalmostanyhumaninconductingasearch［1］。Inthecontextofcomputersearchalgorithms,heuristicimpliessimplesearchthroughspecificknowledge.Un_deraspecialfunction,ateverystepthesealg_orithmssearchthenodewithhighestscoreasthenextnodetobeextended.Themajoreffectofheuristicsearchisthatinsomewayitconstrainsthesearchspace.Theshortestpathalgorithmsbasedonheuristicincludescostingalgorithm［1］，branch_and_boundalgorithm［2，3］，greedyalgorithms［4～11］，hill_climbingalgorithms［1～3］，andA*algorithm［12～14］,etc.　　Amongtheknownshortestpathalgorithms,manyofthemusethegreedystrategiesasthesearchstrategiesandexplorehowtodesigndelicaterunningdatastructuresandsearchingalgorithms,soastoimprovetherunningefficiencyofsequentialshortestpathalgorithmsundertheuniformtimecomplexity.Forthetwobranchesoflosslessshortestpathalgorithms：labelsettingandlabelcorrectingalgorithms，expertshavesetforwardvariousrunningdatastructuresandrelatedliteraturesemergeendlessly.Refs.［15～19］havemadedetailedanalysisandcomparisonfortheshortestpathalgorithmspresentedbefore.Becausetherealnetworks,suchastransportationnetworks,telecommunicationnetworks,facilitynetworksandhydrographicnetworksdonotconcernnegativeweights,afamouslabelsettingalgorithmadoptinggreedystrategy,namelyDijkstra'salgorithm,hasattractedwideattentionandgotwidedisseminationandapplications.Dijkstra'salgorithmisthemostmatureshortestpathalgorithmtheoreticallyuptodate［8,16,20］andhasbecomethecrucialalgorithmforthenetworkanalysismodulesofmanylargeGISplatforms.Differentimplementationmethodshaveformedalargefami

简介：以便解决动态重建和三维的地质的固体的本地更新问题，推理的拓扑学被用于三维的地质的建模。这能在实现3D推进相应的节自动化的水平由构造的地质的稳固的动态重建并且在3D上在拓扑学上推理弯了表面。这个方法成功地在南京城市里被使用了地质的建模和当模特儿的Zijin金矿。结果证明这个方法适应共面的节和非共面的节数据，并且改进3D的效率地质的建模。

简介：

简介：Syllogisticfuzzyreasoningisintroducedintofuzzysystem,andthenewCascadedFuzzySystem(CFS)ispresented.ThethoroughlytheoreticalanalysisandexperimentalresultsshowthatsyllogisticfuzzyreasoningismorerobustthanallotherimplicationinferencesfornoisedataandthatCFShasbetterrobustnessthanconventionalfuzzysystems,whichprovidethesolidfoundationforCFS'spotentialapplicationinfuzzycontrolandmodelingandsoon.

简介：Inthispaper,afuzzyreasoningbasedtemporalerrorconcealmentmethodisproposed.ThebasictemporalerrorconcealmentisimplementedbyestimatingMotionVector(MV)ofthelostMacroBlock(MB)fromitsneighboringMVs.WhichMVisthemostproperoneisevaluatedbysomecriteria.Generally,twocriteriaarewidelyused,namelySideMatchDistortion(SMD)andSumofAbsoluteDifference(SAD)ofcorrespondingMV.However,eachcriterioncouldonlypartlydescribethestatusoflostblock.Toaccomplishthejudgementmoreaccurately,thetwomeasuresareconsideredtogether.ThusarefinedmeasurebasedonfuzzyreasoningisadoptedtobalancetheeffectsofSMDandSAD.TermsSMDandSADareregardedasfuzzyinputandtheterm‘similarity’asoutputtocompletefuzzyreasoning.ResultoffuzzyreasoningrepresentshowthetestedMVissimilartotheoriginalone.Andk-meansclusteringtechniqueisperformedtodefinethemembershipfunctionofinputfuzzysetsadaptively.Accordingtotheexperimentalresults,theconcealmentbasedonnewmeasureachievesbetterperformance.

简介：