人工神经网络试题及答案.doc

《人工神经网络试题及答案.doc》由会员上传分享，免费在线阅读，更多相关内容在应用文档-天天文库。

1、QuestionOne:TheweightupdatingrulesoftheperceptronandKohonenneuralnetworkare_____.QuestionTwo:Thelimitationoftheperceptronisthatitcanonlymodellinearlyseparableclasses.ThedecisionboundaryofRBFis__________linear______________________whereasthedecisionboundar

2、yofFFNNis__________________non-linear___________________________.QuestionThree:TheactivationfunctionoftheneuronofthePerceptron,BPnetworkandRBFnetworkarerespectively________________;________________;______________.QuestionFour:Pleasepresenttheidea,objectiv

3、efunctionoftheBPneuralnetworks(FFNN)andthelearningruleoftheneuronattheoutputlayerofFFNN.Youareencouragedtowritedowntheprocesstoproducethelearningrule.QuestionFive:PleasedescribethesimilarityanddifferencebetweenHopfieldNNandBoltzmannmachine.相同：Bothofthemar

4、esingle-layerinter-connectionNNs.Theybothhavesymmetricweightmatrixwhosediagonalelementsarezeroes.不同：ThenumberoftheneuronsofHopfieldNNisthesameasthenumberofthedimension(K)ofthevectordata.Ontheotherhand,BoltzmannmachinewillhaveK+Lneurons.ThereareLhiddenneur

5、onsBoltzmannmachinehasKneuronsthatservesasbothinputneuronsandoutputneurons(Auto-associationBoltzmannmachine).QuestionSix:Pleaseexplainthetermsintheaboveequationindetail.PleasedescribetheweightupdatingequationsofeachnodeinthefollowingFFNNusingtheBPlearning

6、algorithm.(PPT原题y=φ(net)=φ(w0+w1x1+w2x2))W0=w0+W1=w1+W2=w2+QuestionSeven:PleasetryyourbesttopresentthecharacteristicsofRBFNN.(1)RBFnetworkshaveonesinglehiddenlayer.(2)InRBFtheneuronmodelofthehiddenneuronsisdifferentfromtheoneoftheoutputnodes.(3)Thehiddenl

7、ayerofRBFisnon-linear,theoutputlayerofRBFislinear.(4)TheargumentofactivationfunctionofeachhiddenneuroninaRBFNNcomputestheEuclideandistancebetweeninputvectorandthecenterofthatunit.(5)RBFNNusesGaussianfunctionstoconstructlocalapproximationstonon-linearI/Oma

8、pping.QuestionEight:Generally,theweightvectorsofallneuronsofSOMisadjustedintermsofthefollowingrule:wj(n+1)=wj(n)+η(n)hi(x)(di(x)j)(x(n)-wj(n)).Pleaseexplaineachtermintheaboveformula.:weightvalueofthej-thneuronatiter