神经网络工具箱的使用

天零ヮ冰釋

1. clear
   - q- C) A% M! _$ G' X9 j' E
2. load('trainset.mat');                   %加载训练集数据8 F4 Q) _' o% m8 |  Z2 Q
   
3. load('testset.mat');                    %加载测试集数据, s1 _- ~6 s/ }$ x\" v* L( g
   
4. p1=trainsetdata([1:179],:);/ S  a3 h5 R8 R6 y* O
   
5. p2=trainsetdata([180:300],:);
   & r: ^( X. q. V4 I  M
6. p=[p1;p2]';0 X6 k7 B; K1 M2 k
   
7. pr=minmax(p);! I- Z2 T/ r& D: v
   
8. goal=[ones(1,179),zeros(1,121);zeros(1,179),ones(1,121)];: s. l% U% L4 E; Q* o; W\" F9 Q
   
9. net=newff(pr,[21,160],{'tansig','logsig'}); %隐含层节点和输出层节点0 g: C( ?\" d7 F, r5 s- s# F
   
10.                                             %取值不太确定，请指教* V4 g/ B$ K* u; r1 R
    
11. net.trainParam.show = 10;# \# M' }% i( E  ?
    
12. net.trainParam.lr = 0.05;
    ! Z* ~, D  k\" T  T& S# g
13. net.trainParam.goal = 1e-10;
    2 e/ v4 r& i) F; r* ~\" t. |2 k
14. net.trainParam.epochs = 50000;$ @8 B) @6 p3 r
    
15. net = train(net,p,goal);* H* q4 H# R9 X7 [
    
16. x=testsetdata';) c5 R; P1 W+ j1 }$ N. A  F1 {% N/ S( q
    
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. ' J- i! V3 G& Q+ C- O& [
4. > In obs_use at 184 \, \. i: ^) X- ~2 ^9 @\\" a* r# X0 R
   
6.   In newff>create_network at 127
   
7. 5 V. [! L1 X/ A  y& O6 m
8.   In newff at 102  K5 Q4 u/ F- y$ J! t
   
10.   In test at 9' [+ Q5 d% N/ n* v
    
12.           See help for NEWFF to update calls to the new argument list.
    
13. / A5 u% l( D/ u7 A( N$ j! s
14. 
    
15. + s- t! V\\" t& e$ c\\" l
16. ??? Error using ==> trainlm at 109
    
17. 5 ~/ ]\\" r7 h6 \' i* W6 P8 R: s  l
18. Output data size does not match net.outputs{2}.size.
    
19. : {/ H/ k+ \, N: \- j6 X
20. - R% o4 e) u% V6 q7 j% l
    
22. Error in ==> network.train at 107
    
23. % ~# M. M6 N* D! c
24. [net,tr] = feval(net.trainFcn,net,X,T,Xi,Ai,EW,net.trainParam);. O. {: {9 E2 b$ G; @
    
26. 3 m/ u0 R0 x3 S9 f8 t
    
28. Error in ==> test at 15
    
29. ' Y0 W# ^( G: _( N: b( k
30. net = train(net,p,goal);

Warning: NEWFF used in an obsolete way. <br /> <span style=\\"display:none\\">' J- i! V3 G& Q+ C- O& [</span> > In obs_use at 18<font class=\\"jammer\\">4 \, \. i: ^) X- ~2 ^9 @\\" a* r# X0 R</font><br />   In newff>create_network at 127<br /> <span style=\\"display:none\\">5 V. [! L1 X/ A  y& O6 m</span>   In newff at 102<font class=\\"jammer\\">  K5 Q4 u/ F- y$ J! t</font><br />   In test at 9<font class=\\"jammer\\">' [+ Q5 d% N/ n* v</font><br />           See help for NEWFF to update calls to the new argument list.<br /> <span style=\\"display:none\\">/ A5 u% l( D/ u7 A( N$ j! s</span> <br /> <span style=\\"display:none\\">+ s- t! V\\" t& e$ c\\" l</span> ??? Error using ==> trainlm at 109<br /> <span style=\\"display:none\\">5 ~/ ]\\" r7 h6 \' i* W6 P8 R: s  l</span> Output data size does not match net.outputs{2}.size.<br /> <span style=\\"display:none\\">: {/ H/ k+ \, N: \- j6 X</span> <font class=\\"jammer\\">- R% o4 e) u% V6 q7 j% l</font><br /> Error in ==> network.train at 107<br /> <span style=\\"display:none\\">% ~# M. M6 N* D! c</span> [net,tr] = feval(net.trainFcn,net,X,T,Xi,Ai,EW,net.trainParam);<font class=\\"jammer\\">. O. {: {9 E2 b$ G; @</font><br /> <font class=\\"jammer\\">3 m/ u0 R0 x3 S9 f8 t</font><br /> Error in ==> test at 15<br /> <span style=\\"display:none\\">' Y0 W# ^( G: _( N: b( k</span> net = train(net,p,goal);

jiajinshan110

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