[求助]神经网络模拟问题!!!!!!!!!!
<P> 我在神经网络训练样本之后,不知道该如何去利用这些样本去检验和预测样本植.</P><P>请高手指点一二.</P>
<P>急切等待你的回复,谢谢!</P> <P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left>eural Network object:<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left><p> </p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> architecture:<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left><p> </p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> numInputs: 1<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> numLayers: 2<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> biasConnect: <p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> inputConnect: <p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> layerConnect: <p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> outputConnect: <p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> targetConnect: <p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left><p> </p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> numOutputs: 1(read-only)<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> numTargets: 1(read-only)<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> numInputDelays: 0(read-only)<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> numLayerDelays: 0(read-only)<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left><p> </p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> subobject structures:<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left><p> </p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> inputs: {1x1 cell} of inputs<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> layers: {2x1 cell} of layers<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> outputs: {1x2 cell} containing 1 output<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> targets: {1x2 cell} containing 1 target<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> biases: {2x1 cell} containing 2 biases<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> inputWeights: {2x1 cell} containing 1 input weight<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> layerWeights: {2x2 cell} containing 1 layer weight<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left><p> </p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> functions:<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left><p> </p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> adaptFcn: 'trains'<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> initFcn: 'initlay'<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> performFcn: 'mse'<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> trainFcn: 'trainlm'<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left><p> </p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> parameters:<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left><p> </p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> adaptParam: .passes<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> initParam: (none)<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> performParam: (none)<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> trainParam: .epochs, .goal, .max_fail, .mem_reduc, <p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> .min_grad, .mu, .mu_dec, .mu_inc, <p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> .mu_max, .show, .time, .lr<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left><p> </p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> weight and bias values:<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left><p> </p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> IW: {2x1 cell} containing 1 input weight matrix<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> LW: {2x2 cell} containing 1 layer weight matrix<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> b: {2x1 cell} containing 2 bias vectors</P><p><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left>a=sim(net,p)<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left><p> </p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left>a =<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left><p> </p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left> 0.4618 0.3917 0.3617 0.2927 0.5025 0.3945<p></p></P><P 0cm 0cm 0pt; TEXT-ALIGN: left; mso-layout-grid-align: none" align=left></p> </P> 发放和法规和规范化
页:
[1]