matlab 前向神经网络

注：另参考其他网络 fitnet（拟合），patternnet（模式识别）

feedforwardnet

net=feedforwardnet;

同通用语法

feedforwardnet(hiddenSizes,trainFcn)
hiddenSizes为向量，定义多个隐藏层中神经元的数量
trainFcn训练函数，默认trainlm 

info

 dimensions:
 
         numInputs: 1 %一个输入，（一个并发输入）
         numLayers: 2 %两层，一个隐藏层，一个输出层
        numOutputs: 1 %一个输出层
    %numInputDelays: 0
    %numLayerDelays: 0
 %numFeedbackDelays: 0
 %numWeightElements: 10
        %sampleTime: 1
 
    connections:
 
       biasConnect: [1; 1] %向量，两层，均有偏置
      inputConnect: [1; 0] %向量，仅输入与第一层有连接
      layerConnect: [0 0; 1 0] %矩阵，行为目标层，列为出发层，仅1连接到2层
     outputConnect: [0 1] %向量，决定输出层
 
    subobjects:
 
             input: Equivalent to inputs{1}
            output: Equivalent to outputs{2}
 			
 			%全是对象（结构体） 元胞数组
            inputs: {1x1 cell array of 1 input} 
            layers: {2x1 cell array of 2 layers} 
           outputs: {1x2 cell array of 1 output}
            biases: {2x1 cell array of 2 biases}
      inputWeights: {2x1 cell array of 1 weight}
      layerWeights: {2x2 cell array of 1 weight}
 
    functions:
 
          adaptFcn: 'adaptwb'
        adaptParam: (none)
          derivFcn: 'defaultderiv'
         divideFcn: 'dividerand'
       divideParam: .trainRatio, .valRatio, .testRatio
        divideMode: 'sample'
           initFcn: 'initlay'
        performFcn: 'mse'
      performParam: .regularization, .normalization
          plotFcns: {'plotperform', plottrainstate, ploterrhist,
                    plotregression}
        plotParams: {1x4 cell array of 4 params}
          trainFcn: 'trainlm' %训练函数
        trainParam: .showWindow, .showCommandLine, .show, .epochs,
                    .time, .goal, .min_grad, .max_fail, .mu, .mu_dec,
                    .mu_inc, .mu_max
 
    weight and bias values:
 
                IW: {2x1 cell} containing 1 input weight matrix
                LW: {2x2 cell} containing 1 layer weight matrix
                 b: {2x1 cell} containing 2 bias vectors
 
    methods:
 
             adapt: Learn while in continuous use
         configure: Configure inputs & outputs
            gensim: Generate Simulink model
              init: Initialize weights & biases
           perform: Calculate performance
               sim: Evaluate network outputs given inputs
             train: Train network with examples
              view: View diagram
       unconfigure: Unconfigure inputs & outputs
 
    evaluate:       outputs = net(inputs)

layers

>>net.layers{1}

  name: 'Hidden' %自定义的可改
        dimensions: 10 %与size的区别没懂
       distanceFcn: (none)
     distanceParam: (none)
         distances: []
           initFcn: 'initnw'
       netInputFcn: 'netsum' %层函数，加上偏置
     netInputParam: (none)
         positions: []
             range: [10x2 double]
              size: 10 %默认神经元数，net.layers{i}.size
       topologyFcn: (none)
       transferFcn: 'tansig' %转换函数,net.layers{i}.transferFcn
     transferParam: (none)
          userdata: (your custom info)

layersweight

这与info是对应的layerConnect: [0 0; 1 0] %矩阵，行为目标层，列为出发层，仅1连接到2层

>>net.layerWeights{2,1}

            delays: 0
           initFcn: (none)
      initSettings: .range
             learn: true
          learnFcn: 'learngdm'
        learnParam: .lr, .mc
              size: [0 10]
         weightFcn: 'dotprod' %权重函数
       weightParam: (none)
          userdata: (your custom info)