eval_nn.m

其中提到遺傳學的程式碼與應用提供給次淚相向的研究者參考下載

% eval_nn.m

% Read FIS
nn_flc=readfis('nn1_flc');
disp('==>Load nn1_flc FIS complete!');
% Input patterns
P=[0.8  0.8; 0.8  0; 0.8  -0.8; 0.3  0.3;
   0.3  -0.3; 0    0.8;	0    0; 0    -0.8;
  -0.3  0.3; -0.3  -0.3; -0.8  0.8; -0.8  0;
  -0.8  -0.8];
IN_patterns=P'
% Output targets
Targets=[1 0.55 0 0.35 0 0.55 0 -0.55 0 -0.35 0 -0.55 -1]
% Evaluate FIS
Fuzzy_out=[evalfis(IN_patterns,nn_flc)]'
% Training a FF NN
net = newff([-1 1;-1 1],[4 1],{'tansig' 'purelin'});
net.trainParam.epochs = 1500;
net = train(net,IN_patterns,Targets);
% Display results
Targets=[1 0.55 0 0.35 0 0.55 0 -0.55 0 -0.35 0 -0.55 -1]
Neural_out = sim(net,IN_patterns)
% Plot NN training results
sample=1:13;
figure(2);plot(sample,Targets,sample,Neural_out,'--')
title('NN training results plots(Solid=Targets,dash=NN out)');
xlabel('(Use of eval_nn.m) Samples');
ylabel('Outputs')
% Plot 3-d
gop=input('Take several minutes to plot 3D graph,OK<CR>?(0=NO plot):');
if isempty(gop)
    x=-1:0.02:1; y=x;
    [X, Y]=meshgrid(x,y);
    [M, N]=size(X);
    for I=1:M
    for J=1:N
        F_out(I,J)=evalfis([X(I,J) Y(I,J)],nn_flc);
        N_out(I,J)=sim(net,[X(I,J); Y(I,J)]);
    end
    end
    figure(3);mesh(X,Y,F_out)
    figure(4);mesh(X,Y,N_out)
end    
disp('Done eval_nn<<<');