getting error "Undefined function or method 'logistic' for input arguments of type 'double'.""

Question

vandana 2015-6-29

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/225732-getting-error-undefined-function-or-method-logistic-for-input-arguments-of-type-double

编辑： Walter Roberson 2015-6-30

I have written a code based on neural network and when I ma trying to run code getting error as "Undefined function or method 'logistic' for input arguments of type 'double'" I have given the portion of code where getting error.

double logistic(x)
    if(x > 100.0)
        x = 1.0;
    else if (x < -100.0) 
            x = 0.0;
    else x = 1.0/(1.0+exp(-x));
     x;
        end
    end
function [V_OUT]= feed_forward_signals(MAT_INOUT, V_IN, V_OUT, V_BIAS, size1, size2, layer)
for row = 1:size2
    V_OUT(row)=0;
    for col = 1:size1
     V_OUT(row)=V_OUT(row)+MAT_INOUT(row,col)*V_IN(col);
    end
      V_OUT(row)=V_OUT(row)+V_BIAS(row);
end
      if(layer==0)
          V_OUT(row) = tanh(V_OUT(row));
      end
      if(layer==1) 
          V_OUT(row) = logistic(V_OUT(row));

please help

end

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Greg Heath 2015-6-29

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/225732-getting-error-undefined-function-or-method-logistic-for-input-arguments-of-type-double#answer_184383

在 MATLAB Online 中打开

If you have the Neural Network Toolbox just use

 LOGSIG or TANSIG
 otherwise, just use TANH.

Also see

Subject: NEURAL NET TRAINING SPEED USING A MODIFIED ELLIOT SIGMOID

From: Greg Heath

Date: 14 Oct, 2012 06:18:08

Hope this helps.

Thank you for formally accepting my answer

Greg

2 个评论
显示无隐藏无

vandana 2015-6-30

Hi, thanks for your reply. but does LOGSIG or TANSIG functions will give the same results as logistic. because I am getting good results with logistic function only. TANH is not giving good results.

Greg Heath 2015-6-30

在 MATLAB Online 中打开

 [x,t] = simplefit_dataset;
  z    = [ x; t]; 
 1. TANSIG is MATLAB's version of TANH.
    y1   = tansig(z);
    y2   = tanh(z);
 >> ans = minmax(y1-y2)
    ans = -3.3307e-16   3.3307e-16
          -3.3307e-16   3.3307e-16
 2. LOGSIG is MATLAB's version of the LOGISTIC.  
 3. It has been demonstrated in the comp.ai.neural-nets FAQ that tansig hidden nodes are better than logsig when inputs have zero mean.