神经网络工具箱的使用

天零ヮ冰釋

1. clear
   $ @& u; `# V; M
2. load('trainset.mat');                   %加载训练集数据: i' t* {6 v* @* @' C
   
3. load('testset.mat');                    %加载测试集数据. R# _( ]' M1 v. j+ T# j
   
4. p1=trainsetdata([1:179],:);
   8 `8 t& j7 ^2 G2 s\" j* _
5. p2=trainsetdata([180:300],:);- S2 L, d, F& i% i& k
   
6. p=[p1;p2]';
   2 _  l& {- S' `8 Q# Q
7. pr=minmax(p);( ]3 W& [# Q5 W- v$ f/ @9 @0 v
   
8. goal=[ones(1,179),zeros(1,121);zeros(1,179),ones(1,121)];
   * ]- I* v9 f7 c' C
9. net=newff(pr,[21,160],{'tansig','logsig'}); %隐含层节点和输出层节点
   - J; w  o/ q/ c2 E1 q9 j+ M
10.                                             %取值不太确定，请指教
    ) P  t1 d+ q8 B; E1 ~) j7 `
11. net.trainParam.show = 10;
    7 _! m7 E\" `1 [6 Y0 I) R; Y+ q$ C
12. net.trainParam.lr = 0.05;* V$ y; g- i, I/ D
    
13. net.trainParam.goal = 1e-10;
    5 I\" \7 E% v+ V6 r. y8 Q
14. net.trainParam.epochs = 50000;
    . X( p3 J, u) P: J; }
15. net = train(net,p,goal);: P, q9 y+ m; C; ?
    
16. x=testsetdata';
    $ K4 [. a' ~* Q- c$ D) t8 G3 e4 A
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. & g3 @+ \8 _5 f% ?' g
4. > In obs_use at 18
   
5. 6 n* a\\" r! B0 ^\\" Z# ]/ E# I
6.   In newff>create_network at 127$ i1 p, Y# Q4 {& Q
   
8.   In newff at 102( R- t* f; @/ r6 l; V7 D4 b9 U
   
10.   In test at 9
    
11. * B, ^) u$ A& B* m6 Y' J! g
12.           See help for NEWFF to update calls to the new argument list.- u8 B' W; U* N0 K0 s5 I/ `
    
14. 
    
15.   r; {, a7 o6 U# Q7 {: K8 A2 e7 q
16. ??? Error using ==> trainlm at 1091 m\\" x! `% t0 B: w* P
    
18. Output data size does not match net.outputs{2}.size.
    
19. % i8 g  R) D# h# e+ T/ ?* X  Q8 T2 h
20. 8 r* F. P) m6 L* T
    
22. Error in ==> network.train at 107. I2 u$ x$ n1 ~1 f# @3 ?0 g\\" i
    
24. [net,tr] = feval(net.trainFcn,net,X,T,Xi,Ai,EW,net.trainParam);* T7 J; I' `& ?4 U* r& U
    
26. 1 |/ m7 G# t! Y% v7 B8 u
    
28. Error in ==> test at 15
    
29. - Q, j6 F' d2 V) y' u+ K
30. net = train(net,p,goal);

Warning: NEWFF used in an obsolete way. <br /> <span style=\\"display:none\\">& g3 @+ \8 _5 f% ?' g</span> > In obs_use at 18<br /> <span style=\\"display:none\\">6 n* a\\" r! B0 ^\\" Z# ]/ E# I</span>   In newff>create_network at 127<font class=\\"jammer\\">$ i1 p, Y# Q4 {& Q</font><br />   In newff at 102<font class=\\"jammer\\">( R- t* f; @/ r6 l; V7 D4 b9 U</font><br />   In test at 9<br /> <span style=\\"display:none\\">* B, ^) u$ A& B* m6 Y' J! g</span>           See help for NEWFF to update calls to the new argument list.<font class=\\"jammer\\">- u8 B' W; U* N0 K0 s5 I/ `</font><br /> <br /> <span style=\\"display:none\\">  r; {, a7 o6 U# Q7 {: K8 A2 e7 q</span> ??? Error using ==> trainlm at 109<font class=\\"jammer\\">1 m\\" x! `% t0 B: w* P</font><br /> Output data size does not match net.outputs{2}.size.<br /> <span style=\\"display:none\\">% i8 g  R) D# h# e+ T/ ?* X  Q8 T2 h</span> <font class=\\"jammer\\">8 r* F. P) m6 L* T</font><br /> Error in ==> network.train at 107<font class=\\"jammer\\">. I2 u$ x$ n1 ~1 f# @3 ?0 g\\" i</font><br /> [net,tr] = feval(net.trainFcn,net,X,T,Xi,Ai,EW,net.trainParam);<font class=\\"jammer\\">* T7 J; I' `& ?4 U* r& U</font><br /> <font class=\\"jammer\\">1 |/ m7 G# t! Y% v7 B8 u</font><br /> Error in ==> test at 15<br /> <span style=\\"display:none\\">- Q, j6 F' d2 V) y' u+ K</span> net = train(net,p,goal);

jiajinshan110

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