神经网络工具箱的使用

天零ヮ冰釋

1. clear
     q, w) }8 }. V* m0 A5 z
2. load('trainset.mat');                   %加载训练集数据- k0 f& b5 y! V' J9 g1 e
   
3. load('testset.mat');                    %加载测试集数据( Z7 @8 o- q4 }/ P# Y
   
4. p1=trainsetdata([1:179],:);
   * f: A& V3 H- u1 I4 d
5. p2=trainsetdata([180:300],:);9 D1 [' Y5 U, }
   
6. p=[p1;p2]';0 b6 s. [5 J/ ~% t# g. a  J
   
7. pr=minmax(p);' D! P\" ~+ N& q+ L* X  f
   
8. goal=[ones(1,179),zeros(1,121);zeros(1,179),ones(1,121)];/ N% @; \( V# _& x
   
9. net=newff(pr,[21,160],{'tansig','logsig'}); %隐含层节点和输出层节点
   + C7 V* {- V8 B) u, X
10.                                             %取值不太确定，请指教- T1 @# Q7 @- ^& h; g9 _
    
11. net.trainParam.show = 10;
    5 t  c6 X9 v% S- ~3 y3 ?/ Y& r. @
12. net.trainParam.lr = 0.05;
      j  l, }4 |( d4 D; e8 \
13. net.trainParam.goal = 1e-10;
    ' c' d' k9 y3 L\" ^! s
14. net.trainParam.epochs = 50000;
    : d- L, q% e( l! u
15. net = train(net,p,goal);
    6 |  [1 Z3 C, Y# Z- I$ W
16. x=testsetdata';1 `\" e. l5 X5 D/ @& `( Z
    
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.
   
3. ' }, w8 q6 t* D2 R! s% d
4. > In obs_use at 18
   
5. ( ]' f1 }6 W7 c) ^' W
6.   In newff>create_network at 127* U- w* b$ N8 M! n' U6 @
   
8.   In newff at 102
   
9.   H* X8 b' e- D: N) d5 Z
10.   In test at 9' [1 @  B- K3 T- j+ r0 F) K9 c
    
12.           See help for NEWFF to update calls to the new argument list.4 R) F# M7 u* H1 z5 R% x+ q3 Q  T
    
14. : i+ T3 N! _9 I/ U6 |/ }
    
16. ??? Error using ==> trainlm at 109
    
17. # H8 Q; M/ ^- I) i/ n+ r
18. Output data size does not match net.outputs{2}.size.' E- m! B. p# {, r/ k
    
20. 4 c\\" P* G! Y: G
    
22. Error in ==> network.train at 107+ c  z+ M  s6 ]; {5 i3 U
    
24. [net,tr] = feval(net.trainFcn,net,X,T,Xi,Ai,EW,net.trainParam);
    
25. \\" U& J. ], s\\" K
26. 
    
27. 0 r\\" J5 l( W+ s% S
28. Error in ==> test at 15
    
29. 6 h$ q/ C6 f) Z: \: X
30. net = train(net,p,goal);

Warning: NEWFF used in an obsolete way. <br /> <span style=\\"display:none\\">' }, w8 q6 t* D2 R! s% d</span> > In obs_use at 18<br /> <span style=\\"display:none\\">( ]' f1 }6 W7 c) ^' W</span>   In newff>create_network at 127<font class=\\"jammer\\">* U- w* b$ N8 M! n' U6 @</font><br />   In newff at 102<br /> <span style=\\"display:none\\">  H* X8 b' e- D: N) d5 Z</span>   In test at 9<font class=\\"jammer\\">' [1 @  B- K3 T- j+ r0 F) K9 c</font><br />           See help for NEWFF to update calls to the new argument list.<font class=\\"jammer\\">4 R) F# M7 u* H1 z5 R% x+ q3 Q  T</font><br /> <font class=\\"jammer\\">: i+ T3 N! _9 I/ U6 |/ }</font><br /> ??? Error using ==> trainlm at 109<br /> <span style=\\"display:none\\"># H8 Q; M/ ^- I) i/ n+ r</span> Output data size does not match net.outputs{2}.size.<font class=\\"jammer\\">' E- m! B. p# {, r/ k</font><br /> <font class=\\"jammer\\">4 c\\" P* G! Y: G</font><br /> Error in ==> network.train at 107<font class=\\"jammer\\">+ c  z+ M  s6 ]; {5 i3 U</font><br /> [net,tr] = feval(net.trainFcn,net,X,T,Xi,Ai,EW,net.trainParam);<br /> <span style=\\"display:none\\">\\" U& J. ], s\\" K</span> <br /> <span style=\\"display:none\\">0 r\\" J5 l( W+ s% S</span> Error in ==> test at 15<br /> <span style=\\"display:none\\">6 h$ q/ C6 f) Z: \: X</span> net = train(net,p,goal);

jiajinshan110

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