神经网络工具箱的使用

天零ヮ冰釋

1. clear
   2 v, e0 @8 s4 c, x
2. load('trainset.mat');                   %加载训练集数据# N# s1 y+ u6 o, k, ?
   
3. load('testset.mat');                    %加载测试集数据
   0 U: ^! G3 [5 k- W  j1 k. |
4. p1=trainsetdata([1:179],:);
   0 j- \) Q* F3 A% X
5. p2=trainsetdata([180:300],:);
   5 k5 W1 d1 h, ?! B- ]
6. p=[p1;p2]';0 v$ }/ o- _% ]# T% E4 F9 O
   
7. pr=minmax(p);; P# n  I8 L) u' `* c
   
8. goal=[ones(1,179),zeros(1,121);zeros(1,179),ones(1,121)];1 L4 G/ k, u. e) l6 P
   
9. net=newff(pr,[21,160],{'tansig','logsig'}); %隐含层节点和输出层节点
     s+ C  n, R0 T# S
10.                                             %取值不太确定，请指教
    , l\" B, U, H: h, V
11. net.trainParam.show = 10;% V\" G4 u+ w1 j6 ?* u4 ]* r' |  L
    
12. net.trainParam.lr = 0.05;\" [  l/ Y$ p/ b# u* `* b
    
13. net.trainParam.goal = 1e-10;: h: l) ?# v3 |  {
    
14. net.trainParam.epochs = 50000;8 |- X2 i1 Y: S- q# S; z$ S
    
15. net = train(net,p,goal);5 I6 j9 F2 Y4 y7 t
    
16. x=testsetdata';* U2 b' V! a% c: ]; L& E3 D  Y
    
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. & T& T2 L, ?: A5 @8 x- e3 w- Z
   
4. > In obs_use at 18/ G- _$ r3 N  h( o
   
6.   In newff>create_network at 127- d- |3 |8 l$ M# y' L
   
8.   In newff at 102
   
9. ( \  r: q1 X6 L3 R
10.   In test at 9
    
11. $ E0 D8 K7 L4 r4 V
12.           See help for NEWFF to update calls to the new argument list.5 U; ~/ y9 [3 w5 f
    
14. 
    
15. 9 A$ M, @+ W; W: n8 I- F
16. ??? Error using ==> trainlm at 109\\" }9 }0 m7 z' B( ?# I- M  e) u
    
18. Output data size does not match net.outputs{2}.size.
    
19. ' }9 O$ D3 }\\" `! r\\" g3 h3 q- i
20. 
    
21. % ?; Z* F1 R8 S1 ?( I
22. Error in ==> network.train at 107. P, `% @8 e, I+ _' V- I+ c/ m
    
24. [net,tr] = feval(net.trainFcn,net,X,T,Xi,Ai,EW,net.trainParam);; K8 O3 d: }, q* K3 C
    
26. 1 v, X1 c$ ~+ C\\" }& q
    
28. Error in ==> test at 15
    
29. ( U9 o% m0 L; f8 r! N0 u0 o) m
30. net = train(net,p,goal);

Warning: NEWFF used in an obsolete way. <font class=\\"jammer\\">& T& T2 L, ?: A5 @8 x- e3 w- Z</font><br /> > In obs_use at 18<font class=\\"jammer\\">/ G- _$ r3 N  h( o</font><br />   In newff>create_network at 127<font class=\\"jammer\\">- d- |3 |8 l$ M# y' L</font><br />   In newff at 102<br /> <span style=\\"display:none\\">( \  r: q1 X6 L3 R</span>   In test at 9<br /> <span style=\\"display:none\\">$ E0 D8 K7 L4 r4 V</span>           See help for NEWFF to update calls to the new argument list.<font class=\\"jammer\\">5 U; ~/ y9 [3 w5 f</font><br /> <br /> <span style=\\"display:none\\">9 A$ M, @+ W; W: n8 I- F</span> ??? Error using ==> trainlm at 109<font class=\\"jammer\\">\\" }9 }0 m7 z' B( ?# I- M  e) u</font><br /> Output data size does not match net.outputs{2}.size.<br /> <span style=\\"display:none\\">' }9 O$ D3 }\\" `! r\\" g3 h3 q- i</span> <br /> <span style=\\"display:none\\">% ?; Z* F1 R8 S1 ?( I</span> Error in ==> network.train at 107<font class=\\"jammer\\">. P, `% @8 e, I+ _' V- I+ c/ m</font><br /> [net,tr] = feval(net.trainFcn,net,X,T,Xi,Ai,EW,net.trainParam);<font class=\\"jammer\\">; K8 O3 d: }, q* K3 C</font><br /> <font class=\\"jammer\\">1 v, X1 c$ ~+ C\\" }& q</font><br /> Error in ==> test at 15<br /> <span style=\\"display:none\\">( U9 o% m0 L; f8 r! N0 u0 o) m</span> net = train(net,p,goal);

jiajinshan110

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