離散Hopfield神經(jīng)網(wǎng)絡(luò)及其在聯(lián)想記憶方面的應(yīng)用.pdf

1、大連理工大學(xué)碩士學(xué)位論文離散Hopfield神經(jīng)網(wǎng)絡(luò)及其在聯(lián)想記憶方面的應(yīng)用姓名：劉照升申請(qǐng)學(xué)位級(jí)別：碩士專業(yè)：運(yùn)籌學(xué)與控制論指導(dǎo)教師：郭崇慧20081201離散Hoptield神經(jīng)網(wǎng)絡(luò)及其在聯(lián)想記憶方面的應(yīng)用DiscreteHopfieldNeuralNetworkanditsApplicationInAssociativeMemo噦AbstractHopfieldneuralnetworkiSaimportantartificial

2、neuralnetworkandiSwidelyusedInthisdissertation，sometheoriesondiscreteHopfieldneuralnetworkanditsapplicationinassociativememoryareinvestigatedindetailAtthesanletime，discreteHopfieldneuralnetworkwithdelayisalsostudiedFist，

3、thestabletheoryondiscreteHopfieldneuralnetworkisinvestigated，theoriginalstabilityresultsareinvestigatedindetail，thestepsrequiredforconvergencehavebeenstudiedindetailaftertheenergyfunctionreachedtheextremism，itisaveryimpo

4、rtantproblemthatisnotconsideredindetailinoriginalproofTheconditionwhichthenon—symmetricalweightmatrixshouldsatisfyforthenetworktoconvergetoastablepointisobtained，thiswillhelpUStounderstandthenetworktoapointTheselectionof

5、theactivationfunctionofaneuronisstudied，wepresentourpointaboutthisquestion，theconvergencepropertyattheaspectofconvergencestepsisimprovedWealsogivesomeexamplescorrespondingeachstabilityresultSOthatwecaildeeplyunderstandth

6、econvergenceresultsandtheconceptoftheoperatingmodes。Secondly，therelatedproblemsofassociativememoryarestudied，theyarethekeyproblemsweinvestigatedFirst，weexplaintheconceptoftheassociativememory，discussthecriteriatoassessth

7、eperformanceoftheassociativememorynetworkandtheperformanceofthemethodstoconstructtheassociativememorynetworkThentheideaandperfofinanceoftheneuralnetworkbasedontheouterproductequationlearningrulealereviewedindetailBasedon

8、theformerresult，weprovethattheneuralnetworkbasedontheouterproductequationleamingruleconvergesafteronlyoneiterationundertheparalleloperationmodeiftheactivatefunctionistakenasweprefer，thisreducesthecomplexityofresearchCont

9、raposetheshortcomingofHopfieldassociativememorynetwork，aprojectbasedontheouterproductequationrulewhichCandiscriminatethememorystateandthespuriousstateiSpresentedAccordingtoOurproject，thenetworkentersa2cycleafteroneiterat

10、ionwheneveraninputiSnotwithinthebasinofattractionofsomememorystatesSomeexamplesconfirmtheefficiencyofourprojectNowadaysfewpeopleconsidertheaffectionofthethresholdintheresearchofassociativememory，thethresholdiSalwaystaken