神经网络工具箱的使用

天零ヮ冰釋

1. clear
   1 B9 u6 x) m+ U1 N( J
2. load('trainset.mat');                   %加载训练集数据
   0 K1 c5 b3 ^% W1 X: z# {
3. load('testset.mat');                    %加载测试集数据
   . r/ B& t9 E; L/ ~7 w1 \
4. p1=trainsetdata([1:179],:);
   ) ^7 w9 P, s6 M  [2 ?$ k+ B5 p
5. p2=trainsetdata([180:300],:);
   , `4 b' J9 V' i, x+ y. M( r
6. p=[p1;p2]';7 X. ^$ d5 [. @: t
   
7. pr=minmax(p);  V( l8 A1 d; V( l! `; [$ C
   
8. goal=[ones(1,179),zeros(1,121);zeros(1,179),ones(1,121)];
   8 q* ]' X3 b$ n% g% ?\" B0 ]
9. net=newff(pr,[21,160],{'tansig','logsig'}); %隐含层节点和输出层节点8 q+ E: e$ O+ L7 @& S! N+ u
   
10.                                             %取值不太确定，请指教1 Q/ y9 P2 t0 ]; h1 B7 U
    
11. net.trainParam.show = 10;
    - H. J  D* l. [. m  w! l
12. net.trainParam.lr = 0.05;4 p- p! Z' u# d  @) t0 b
    
13. net.trainParam.goal = 1e-10;
    $ W& ~( E8 A) b- p0 J1 r) J
14. net.trainParam.epochs = 50000;7 o  z\" q- d& s  a6 q\" R
    
15. net = train(net,p,goal);/ U0 Q6 _  y6 d5 ?; `
    
16. x=testsetdata';
    ( N  I. `* _+ X0 I* l
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. , X- b9 S& s6 B5 z
   
4. > In obs_use at 18
   
5. * m3 K$ D\\" P: |# X& ]0 B# [0 c# ~
6.   In newff>create_network at 127\\" B7 M% n$ ^9 i; x$ W4 U5 ]
   
8.   In newff at 102
   
9. ) d; q) e8 U0 G! x7 l
10.   In test at 9
    
11. 3 a! f1 \8 ?7 U5 k/ _; ?
12.           See help for NEWFF to update calls to the new argument list.
    
13. 9 N1 \$ v: @0 k& z\\" ?
14. 
    
15. , J3 M; J5 G: L* t
16. ??? Error using ==> trainlm at 109
    
17. 9 h* }! U\\" R4 J\\" R5 x
18. Output data size does not match net.outputs{2}.size.
    
19. & `/ R# }+ S' f6 |, f2 @; d+ ^7 ]8 i
20. & t+ n# n\\" c2 l8 a$ t) g
    
22. Error in ==> network.train at 1070 ^* z5 \9 `. m% @/ q$ X
    
24. [net,tr] = feval(net.trainFcn,net,X,T,Xi,Ai,EW,net.trainParam);7 d\\" k4 E3 L1 i. F. x) @0 H
    
26. 2 L* d. B, P7 d1 U4 k) Y
    
28. Error in ==> test at 15
    
29. 7 j- }+ y8 L9 n8 D; q
30. net = train(net,p,goal);

Warning: NEWFF used in an obsolete way. <font class=\\"jammer\\">, X- b9 S& s6 B5 z</font><br /> > In obs_use at 18<br /> <span style=\\"display:none\\">* m3 K$ D\\" P: |# X& ]0 B# [0 c# ~</span>   In newff>create_network at 127<font class=\\"jammer\\">\\" B7 M% n$ ^9 i; x$ W4 U5 ]</font><br />   In newff at 102<br /> <span style=\\"display:none\\">) d; q) e8 U0 G! x7 l</span>   In test at 9<br /> <span style=\\"display:none\\">3 a! f1 \8 ?7 U5 k/ _; ?</span>           See help for NEWFF to update calls to the new argument list.<br /> <span style=\\"display:none\\">9 N1 \$ v: @0 k& z\\" ?</span> <br /> <span style=\\"display:none\\">, J3 M; J5 G: L* t</span> ??? Error using ==> trainlm at 109<br /> <span style=\\"display:none\\">9 h* }! U\\" R4 J\\" R5 x</span> Output data size does not match net.outputs{2}.size.<br /> <span style=\\"display:none\\">& `/ R# }+ S' f6 |, f2 @; d+ ^7 ]8 i</span> <font class=\\"jammer\\">& t+ n# n\\" c2 l8 a$ t) g</font><br /> Error in ==> network.train at 107<font class=\\"jammer\\">0 ^* z5 \9 `. m% @/ q$ X</font><br /> [net,tr] = feval(net.trainFcn,net,X,T,Xi,Ai,EW,net.trainParam);<font class=\\"jammer\\">7 d\\" k4 E3 L1 i. F. x) @0 H</font><br /> <font class=\\"jammer\\">2 L* d. B, P7 d1 U4 k) Y</font><br /> Error in ==> test at 15<br /> <span style=\\"display:none\\">7 j- }+ y8 L9 n8 D; q</span> net = train(net,p,goal);

jiajinshan110

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