untitled.m

lvq神经网络的编程

P=[0.045036 0.108780 0.128584 0.009097 0.000000;
   0.119754 0.646052 0.279757 0.176645 0.030556;
   0.017400 0.009939 0.005213 0.023694 0.727778;
   0.148414 0.657096 0.384014 0.182991 0.000000;
   0.035824 0.267808 0.693310 0.025809 0.005556;
   0.078813 0.828272 0.357950 0.582822 1.000000;
   0.031218 0.020431 0.008688 0.029194 0.944444;
   0.058342 1.000000 1.000000 1.000000 0.111111;
   0.631525 0.309221 0.184188 0.053945 0.694444; 
   0.662231 0.276091 0.288445 0.041252 0.127778;
   0.160696 0.701270 0.279757 0.202031 0.000000;
   0.006141 0.018222 0.065161 0.007193 0.010000;
   1.000000 0.386527 0.496959 0.024328 0.022222;
   0.000000 0.675870 0.707211 0.818913 0.150000;
   0.018731 0.036996 0.004334 0.002962 0.000000]';
C=[1 5 3 4 4 5 3 5 3 2 4 1 2 5 4];
T=ind2vec(C);
plotvec(P,C,'*r');
axis([-3 3 -3 3]);
hold on;
net=newlvq(minmax(P),20,[0.13 0.13 0.22 0.26 0.26],0.1);
w1=net.IW{1};
plot(w1(1,1),w1(1,2),'+r');
axis([-8 8 -3 3]);
hold on;
net.trainParam.epochs=500;
net=train(net,P,T);
plotvec(P,C,'*r');
hold on;
plotvec(net.IW{1}',vec2ind(net.LW{2}),'+');
axis([-8 8 -3 3]);
hold on;
p=[0.01392 0.0254 0.003475 0.002327 0;
 1.000000 0.386527 0.496959 0.024328 0.022222;
 0.018731 0.036996 0.004334 0.002962 0.000000;
 0.058342 1.000000 1.000000 1.000000 0.111111]';
y=sim(net,p);
yc=vec2ind(y);