神经网络工具箱的使用

天零ヮ冰釋

1. clear( K/ T* P4 q4 _7 O
   
2. load('trainset.mat');                   %加载训练集数据' h9 T+ q- ?8 ]\" y4 B& p/ O  o
   
3. load('testset.mat');                    %加载测试集数据5 F4 ?' ?' h2 P4 a' ?\" u0 H
   
4. p1=trainsetdata([1:179],:);+ E: T- q. x6 h0 ]2 l7 B
   
5. p2=trainsetdata([180:300],:);* Z! ?/ L( @  E' _+ F* M% m
   
6. p=[p1;p2]';
   ; b/ I1 H\" H! c( P; y7 P
7. pr=minmax(p);
   ; r$ D' [+ e# R% o0 p  Y$ d
8. goal=[ones(1,179),zeros(1,121);zeros(1,179),ones(1,121)];
   1 N+ [( \/ ~( Y+ T- z% I! r* T
9. net=newff(pr,[21,160],{'tansig','logsig'}); %隐含层节点和输出层节点' a, F1 F3 \' N6 S
   
10.                                             %取值不太确定，请指教- ~$ H5 h# z8 D0 R- O7 W7 ]
    
11. net.trainParam.show = 10;
    # R\" b, @5 X$ ~: C) \! G7 ]- |& `
12. net.trainParam.lr = 0.05;
    - c- S; k/ \. f# l
13. net.trainParam.goal = 1e-10;\" {& a/ l5 A/ I/ x\" A5 c( ]
    
14. net.trainParam.epochs = 50000;
      @/ p) s& U2 l8 ]- W
15. net = train(net,p,goal);- b) R* T: X( P
    
16. x=testsetdata';1 Y; I# t6 o+ v# S5 ~
    
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. 3 ~! h$ A5 ]* r, E& D
4. > In obs_use at 18' [& Y7 K2 l% Q3 i( ?
   
6.   In newff>create_network at 1279 _+ J, O) `7 T
   
8.   In newff at 102
   
9. / n0 i0 o/ e$ K4 {+ ^! V
10.   In test at 9
    
11. 7 v, G0 H1 O* n  X# G% h
12.           See help for NEWFF to update calls to the new argument list.
    
13. \\" y; o6 N: f. f0 d4 ~$ B
14. ; N- G  s1 Q! }3 R$ a, J* c' F7 |
    
16. ??? Error using ==> trainlm at 109
    
17. ' H0 p2 w: W; m3 ]3 B
18. Output data size does not match net.outputs{2}.size.& v6 P7 i, n' [6 o) F0 o0 c
    
20. \\" q+ W, h, u6 @\\" H# f' s2 L
    
22. Error in ==> network.train at 107: a) [$ `* X7 I5 p; L2 h( Y
    
24. [net,tr] = feval(net.trainFcn,net,X,T,Xi,Ai,EW,net.trainParam);
    
25. 6 T3 q! v3 q# t! F7 @\\" g
26. * X; x! u& d5 h; _\\" F
    
28. Error in ==> test at 15
    
29. 5 }2 y. [+ O- ^, z' i/ @
30. net = train(net,p,goal);

Warning: NEWFF used in an obsolete way. <br /> <span style=\\"display:none\\">3 ~! h$ A5 ]* r, E& D</span> > In obs_use at 18<font class=\\"jammer\\">' [& Y7 K2 l% Q3 i( ?</font><br />   In newff>create_network at 127<font class=\\"jammer\\">9 _+ J, O) `7 T</font><br />   In newff at 102<br /> <span style=\\"display:none\\">/ n0 i0 o/ e$ K4 {+ ^! V</span>   In test at 9<br /> <span style=\\"display:none\\">7 v, G0 H1 O* n  X# G% h</span>           See help for NEWFF to update calls to the new argument list.<br /> <span style=\\"display:none\\">\\" y; o6 N: f. f0 d4 ~$ B</span> <font class=\\"jammer\\">; N- G  s1 Q! }3 R$ a, J* c' F7 |</font><br /> ??? Error using ==> trainlm at 109<br /> <span style=\\"display:none\\">' H0 p2 w: W; m3 ]3 B</span> Output data size does not match net.outputs{2}.size.<font class=\\"jammer\\">& v6 P7 i, n' [6 o) F0 o0 c</font><br /> <font class=\\"jammer\\">\\" q+ W, h, u6 @\\" H# f' s2 L</font><br /> Error in ==> network.train at 107<font class=\\"jammer\\">: a) [$ `* X7 I5 p; L2 h( Y</font><br /> [net,tr] = feval(net.trainFcn,net,X,T,Xi,Ai,EW,net.trainParam);<br /> <span style=\\"display:none\\">6 T3 q! v3 q# t! F7 @\\" g</span> <font class=\\"jammer\\">* X; x! u& d5 h; _\\" F</font><br /> Error in ==> test at 15<br /> <span style=\\"display:none\\">5 }2 y. [+ O- ^, z' i/ @</span> net = train(net,p,goal);

jiajinshan110

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