神经网络工具箱的使用

天零ヮ冰釋

1. clear7 t3 R\" w% B) A2 Y* }\" y
   
2. load('trainset.mat');                   %加载训练集数据: c5 V  n/ R: s9 _# X* c* U
   
3. load('testset.mat');                    %加载测试集数据2 W; }- D\" x& p
   
4. p1=trainsetdata([1:179],:);8 J' Q  _) d- Y\" H
   
5. p2=trainsetdata([180:300],:);' u& y\" `  D( o0 F; B: i
   
6. p=[p1;p2]';
   4 C: T! a& W4 ~3 j
7. pr=minmax(p);5 B/ R: d9 O2 A
   
8. goal=[ones(1,179),zeros(1,121);zeros(1,179),ones(1,121)];3 a( J  ~+ S0 E0 W; T7 V, ?
   
9. net=newff(pr,[21,160],{'tansig','logsig'}); %隐含层节点和输出层节点
   * K\" |) Z8 D5 s$ n4 y0 O9 o( y
10.                                             %取值不太确定，请指教; O# P- d, r* R. n! w
    
11. net.trainParam.show = 10;% S  ]3 q\" B. i$ u$ c
    
12. net.trainParam.lr = 0.05;$ g, }8 h3 Z1 ^$ D  C
    
13. net.trainParam.goal = 1e-10;
    & e4 a/ U( z* V3 B2 C0 p' z/ P
14. net.trainParam.epochs = 50000;! j0 O) c/ w- A+ n6 {
    
15. net = train(net,p,goal);
    - X4 p) U  c( r5 ?6 N
16. x=testsetdata';
    % U9 t5 Z9 ~9 ^! Z* t2 O
17. y0=sim(net,p)

复制代码

下面是运行结果，请各位高手帮忙改改啊，主要是节点数那块儿不知道该怎么取值

2. Warning: NEWFF used in an obsolete way. & A) c3 |$ b3 |; c0 J\\" z* ^+ b
   
4. > In obs_use at 18
   
5. 8 ^! f* _. z/ [\\" y# [7 W9 a; l
6.   In newff>create_network at 127; Q2 {; b! G. n6 _
   
8.   In newff at 1028 p6 r4 C4 {4 Y
   
10.   In test at 9- K* P1 q  R5 b
    
12.           See help for NEWFF to update calls to the new argument list.
    
13. ! F. h9 v* S8 k
14. / m7 E0 a2 H( l2 I% l! F
    
16. ??? Error using ==> trainlm at 109
    
17. 4 H( `+ U8 c1 o! _( X
18. Output data size does not match net.outputs{2}.size.
    
19. 3 g- C9 c; D# H, J$ o
20. 0 d% H/ U: H0 W) p5 o' d
    
22. Error in ==> network.train at 107& r5 V- l$ k0 w1 {0 `% m
    
24. [net,tr] = feval(net.trainFcn,net,X,T,Xi,Ai,EW,net.trainParam);: g\\" \/ D  q) s% W
    
26. 
    
27. 4 [9 `/ R. H2 q6 j) @4 W5 L$ u  O
28. Error in ==> test at 158 V6 F. |\\" o8 f0 J9 t
    
30. net = train(net,p,goal);

Warning: NEWFF used in an obsolete way. <font class=\\"jammer\\">& A) c3 |$ b3 |; c0 J\\" z* ^+ b</font><br /> > In obs_use at 18<br /> <span style=\\"display:none\\">8 ^! f* _. z/ [\\" y# [7 W9 a; l</span>   In newff>create_network at 127<font class=\\"jammer\\">; Q2 {; b! G. n6 _</font><br />   In newff at 102<font class=\\"jammer\\">8 p6 r4 C4 {4 Y</font><br />   In test at 9<font class=\\"jammer\\">- K* P1 q  R5 b</font><br />           See help for NEWFF to update calls to the new argument list.<br /> <span style=\\"display:none\\">! F. h9 v* S8 k</span> <font class=\\"jammer\\">/ m7 E0 a2 H( l2 I% l! F</font><br /> ??? Error using ==> trainlm at 109<br /> <span style=\\"display:none\\">4 H( `+ U8 c1 o! _( X</span> Output data size does not match net.outputs{2}.size.<br /> <span style=\\"display:none\\">3 g- C9 c; D# H, J$ o</span> <font class=\\"jammer\\">0 d% H/ U: H0 W) p5 o' d</font><br /> Error in ==> network.train at 107<font class=\\"jammer\\">& r5 V- l$ k0 w1 {0 `% m</font><br /> [net,tr] = feval(net.trainFcn,net,X,T,Xi,Ai,EW,net.trainParam);<font class=\\"jammer\\">: g\\" \/ D  q) s% W</font><br /> <br /> <span style=\\"display:none\\">4 [9 `/ R. H2 q6 j) @4 W5 L$ u  O</span> Error in ==> test at 15<font class=\\"jammer\\">8 V6 F. |\\" o8 f0 J9 t</font><br /> net = train(net,p,goal);

jiajinshan110

..................