神经网络工具箱的使用

天零ヮ冰釋

1. clear
   ) K( d7 F* o% M, ~! s
2. load('trainset.mat');                   %加载训练集数据9 Q- R7 B$ e4 P* J2 ~3 B( y
   
3. load('testset.mat');                    %加载测试集数据
   ; O5 S+ N9 ^! K\" ?: Y4 ~7 U
4. p1=trainsetdata([1:179],:);
   9 J6 K% k- K; [% h# G( r. v5 }
5. p2=trainsetdata([180:300],:);
   7 ~/ b( R- A+ a: B* @3 {# }& I
6. p=[p1;p2]';
   \" }% V) o2 c9 z/ M# Z4 @3 \
7. pr=minmax(p);
   4 t2 Y6 S1 N( b6 v8 @* j
8. goal=[ones(1,179),zeros(1,121);zeros(1,179),ones(1,121)];) x* M+ [4 G4 H. L0 t
   
9. net=newff(pr,[21,160],{'tansig','logsig'}); %隐含层节点和输出层节点( [\" m7 n1 k4 l- B* z( [( B9 V
   
10.                                             %取值不太确定，请指教
    6 \: S+ B( s3 T- W: [! i. B
11. net.trainParam.show = 10;
    & v( J# q0 z! v
12. net.trainParam.lr = 0.05;+ f: Q\" F+ c5 O, O
    
13. net.trainParam.goal = 1e-10;/ [9 W\" P- B2 J# G9 C
    
14. net.trainParam.epochs = 50000;
    2 h0 W( {( u9 ~! S% s; k
15. net = train(net,p,goal);
    & a/ A/ P7 ~- n7 g# _7 u
16. x=testsetdata';
    # N9 Q: }0 _% T2 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.
   
3. # z7 V0 J$ e% ?. z' J/ i0 W* i6 [; T; J
4. > In obs_use at 18. ]; L* N2 M; `  e. e$ s' q
   
6.   In newff>create_network at 1277 K. p1 g! a0 e: F\\" L, g: I
   
8.   In newff at 1021 d8 o7 h9 [/ M- z8 J% S# L& K
   
10.   In test at 9\\" i) j# w  n8 c
    
12.           See help for NEWFF to update calls to the new argument list.9 ?4 C9 ~2 j/ F. |- h2 p
    
14. 
    
15. : Y\\" O: f/ R0 ]+ g, `* b7 w4 _0 ~
16. ??? Error using ==> trainlm at 109' z9 F+ j8 R2 q7 n
    
18. Output data size does not match net.outputs{2}.size.8 D. v\\" }8 k& u+ C
    
20. + j1 d4 A* y0 a8 q+ z  k
    
22. Error in ==> network.train at 107
    
23. ) g% S- x/ y# V* v3 A
24. [net,tr] = feval(net.trainFcn,net,X,T,Xi,Ai,EW,net.trainParam);1 `) f- p! z9 X/ l) u
    
26. 1 T& j+ W; ^0 U% _) G) ]- g
    
28. Error in ==> test at 15+ L  ~/ Q: c4 f; D; ~
    
30. net = train(net,p,goal);

Warning: NEWFF used in an obsolete way. <br /> <span style=\\"display:none\\"># z7 V0 J$ e% ?. z' J/ i0 W* i6 [; T; J</span> > In obs_use at 18<font class=\\"jammer\\">. ]; L* N2 M; `  e. e$ s' q</font><br />   In newff>create_network at 127<font class=\\"jammer\\">7 K. p1 g! a0 e: F\\" L, g: I</font><br />   In newff at 102<font class=\\"jammer\\">1 d8 o7 h9 [/ M- z8 J% S# L& K</font><br />   In test at 9<font class=\\"jammer\\">\\" i) j# w  n8 c</font><br />           See help for NEWFF to update calls to the new argument list.<font class=\\"jammer\\">9 ?4 C9 ~2 j/ F. |- h2 p</font><br /> <br /> <span style=\\"display:none\\">: Y\\" O: f/ R0 ]+ g, `* b7 w4 _0 ~</span> ??? Error using ==> trainlm at 109<font class=\\"jammer\\">' z9 F+ j8 R2 q7 n</font><br /> Output data size does not match net.outputs{2}.size.<font class=\\"jammer\\">8 D. v\\" }8 k& u+ C</font><br /> <font class=\\"jammer\\">+ j1 d4 A* y0 a8 q+ z  k</font><br /> Error in ==> network.train at 107<br /> <span style=\\"display:none\\">) g% S- x/ y# V* v3 A</span> [net,tr] = feval(net.trainFcn,net,X,T,Xi,Ai,EW,net.trainParam);<font class=\\"jammer\\">1 `) f- p! z9 X/ l) u</font><br /> <font class=\\"jammer\\">1 T& j+ W; ^0 U% _) G) ]- g</font><br /> Error in ==> test at 15<font class=\\"jammer\\">+ L  ~/ Q: c4 f; D; ~</font><br /> net = train(net,p,goal);

jiajinshan110

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