多重指數(shù)加權(quán)雙向聯(lián)想記憶模型及其在多證據(jù)推理中的決策性能研究.pdf

1、南京航空航天大學(xué)碩士學(xué)位論文多重指數(shù)加權(quán)雙向聯(lián)想記憶模型及其在多證據(jù)推理中的決策性能研究姓名：蔡駿申請(qǐng)學(xué)位級(jí)別：碩士專(zhuān)業(yè)：計(jì)算機(jī)軟件與理論指導(dǎo)教師：陳松燦20000201多重指數(shù)加權(quán)雙向聯(lián)想記憶模型及其在多證據(jù)推理中的決策性能研究AbstractInthispaper，amethodformodelingthelearningofheliefcombinationinevidencereasoningusinganeuralnetwor

2、kispresentedAcentralizednetworkcomposedofmultipleexponentialbidirectionalassociativememoriessharingasingleoutputarrayofneuronsisdesignedtoprocesstheuncertaintymanagementofmanypiecesofevidencesimultaneouslYThestabilityoft

3、heproposedmultipleexponentialbidirectionalassociativememories(muIti—eBAM)，multipleexponentialmodifiedbidirectionalassociativememories(multiMeBAM)andmultipleImprovedexponentialbidirectionalassociativememories(multi—IeBAM)

4、networkisprovedAmajorityrtlleofdecisionmakinginpresentationofmultipleevidenceisalsofoundbythestudyofsignal—noiseratioofmultiplenetworks，whichensuresthattheultimateoutputofthenetworkisdeterminedbyoutputwhichhasthehighests

5、upportrateAcentralizednetworkcomposedofmulti—valuemultipleweightedexponentialbidirectionalassociativememories(muiti—MWIeBAM)sharingasingleoutputarrayofneuronsisalsoproposedItsstabi1ity，insynchronousandasynchronousmodes，i

6、sprovedbydefininganenergyfunctionwhichdecreaseswithchangesofneuronstatesThetheoreticalanalysisandcomputersimulationsindicatethatmultipleexponentialbidirectionalassociativememoryhashigherstoragecapabilityandstrongererrorc

7、orrectingability，whichgrantsthatexperts(singlenetwork)couldstil1reachtherightdecisionunderacertaindegreeofdisturbanceFinallythegivenresultsofsimulationarecoincidentwiththeintuitionofhumanreasoningKeywords：DecisionMaking，