附MATLAB 神经网络工具箱

附MATLAB 神经网络工具箱
Graphical user interface functions.
    nctool      - Neural network classification tool.
    nftool      - Neural network fitting tool.
    nprtool     - Neural Network pattern recognition tool.
    nntool      - Neural Network Toolbox graphical user interface.
    nntraintool - Neural network training tool.
    view        - View a neural network.

  Analysis functions.
    confusion - Classification confusion matrix.
    errsurf   - Error surface of single input neuron.
    maxlinlr  - Maximum learning rate for a linear layer.
    roc       - Receiver operating characteristic.

  Distance functions.
    boxdist  - Box distance function.
    dist     - Euclidean distance weight function.
    mandist  - Manhattan distance weight function.
    linkdist - Link distance function.

  Formatting data.
    combvec  - Create all combinations of vectors.
    con2seq  - Convert concurrent vectors to sequential vectors.
    concur   - Create concurrent bias vectors.
    dividevec  - Create all combinations of vectors.
    ind2vec  - Convert indices to vectors.
    minmax   - Ranges of matrix rows.
    nncopy   - Copy matrix or cell array.
    normc    - Normalize columns of a matrix.
    normr    - Normalize rows of a matrix.
    pnormc   - Pseudo-normalize columns of a matrix.
    quant    - Discretize values as multiples of a quantity.
    seq2con  - Convert sequential vectors to concurrent vectors.
    vec2ind  - Convert vectors to indices.

  Initialize network functions.
    initlay  - Layer-by-layer network initialization function.

  Initialize layer functions.
    initnw   - Nguyen-Widrow layer initialization function.
    initwb   - By-weight-and-bias layer initialization function.

  Initialize weight and bias functions.
    initcon  - Conscience bias initialization function.
    initzero - Zero weight/bias initialization function.
    initsompc - Initialize SOM weights with principle components.
    midpoint - Midpoint weight initialization function.
    randnc   - Normalized column weight initialization function.
    randnr   - Normalized row weight initialization function.
    rands    - Symmetric random weight/bias initialization function.

  Learning functions.
    learncon  - Conscience bias learning function.
    learngd   - Gradient descent weight/bias learning function.
    learngdm  - Gradient descent w/momentum weight/bias learning function.
    learnh    - Hebb weight learning function.
    learnhd   - Hebb with decay weight learning function.
    learnis   - Instar weight learning function.
    learnk    - Kohonen weight learning function.
    learnlv1  - LVQ1 weight learning function.
    learnlv2  - LVQ2 weight learning function.
    learnos   - Outstar weight learning function.
    learnsomb - Batch self-organizing map weight learning function.
    learnp    - Perceptron weight/bias learning function.
    learnpn   - Normalized perceptron weight/bias learning function.
    learnsom  - Self-organizing map weight learning function.
    learnwh   - Widrow-Hoff weight/bias learning rule.

  Line search functions.
    srchbac  - Backtracking search.
    srchbre  - Brent's combination golden section/quadratic interpolation.
    srchcha  - Charalambous' cubic interpolation.
    srchgol  - Golden section search.
    srchhyb  - Hybrid bisection/cubic search.

  Net input functions.
    netprod  - Product net input function.
    netsum   - Sum net input function.

  Network creation functions.
    network  - Create a custom neural network.
    newc     - Create a competitive layer.
    newcf    - Create a cascade-forward backpropagation network.
    newdtdnn - Create a distributed time delay neural network.
    newelm   - Create an Elman backpropagation network.
    newfit   - Createa a fitting network.
    newff    - Create a feed-forward backpropagation network.
    newfftd  - Create a feed-forward input-delay backprop network.
    newgrnn  - Design a generalized regression neural network.
    newhop   - Create a Hopfield recurrent network.
    newlin   - Create a linear layer.
    newlind  - Design a linear layer.
    newlvq   - Create a learning vector quantization network.
    newnarx   - Create a feed-forward backpropagation network with feedback
      from output to input.
    newnarxsp   - Create an NARX network in series-parallel arrangement.
    newp     - Create a perceptron.
    newpnn   - Design a probabilistic neural network.
    newpr    - Create a pattern recognition network.
    newrb    - Design a radial basis network.
    newrbe   - Design an exact radial basis network.
    newsom   - Create a self-organizing map.

转载于:https://www.cnblogs.com/zyx2007/archive/2012/09/21/2697020.html