大师作文网BP神经网络数据分类总出错.
来源：学生作业帮 编辑：大师作文网作业帮 分类：数学作业 时间：2024/11/11 12:12:41
BP神经网络数据分类总出错.
%构建训练样本中的输入向量p
p1=[13 44.8 24.7 0.97;
14 42.3 24.4 0.95;
15 42.5 24.6 0.97;
17 45.6 25.8 0.96];
p2=[21 37.6 21.3 0.92;
25 37.6 21.7 0.90;
23 35.4 20.8 0.94;
24 37.2 21.6 0.87];
p3=[26 32.3 20.2 0.84;
27 27.6 18.7 0.81;
26 33.7 20.8 0.83;
29 32.3 20.3 0.83];
p4=[30 21.5 16.4 0.77;
32 20.6 15.8 0.76;
36 22.6 16.8 0.77;
35 21.3 16.2 0.75];
p=[p1 p2 p3 p4];
%构建训练样本中的目标向量t
t1=[1 1;1 1;1 1;1 1];
t2=[1 0;1 0;1 0;1 0];
t3=[0 1;0 1;0 1;0 1];
t4=[0 0;0 0;0 0;0 0];
t=[t1 t2 t3 t4];
p=p';t=t';
net=newff(p,t,3,{'tansig','logsig'},'traingd');
net.trainParam.epochs=500;
net.trainParam.goal=0.01;
net=train(net,p,t);
%测试样本进行分类
p_test=[13 44.7 24.3 0.95;
12 41.2 23.7 0.96;
13 42.5 24.6 0.97;
16 40.5 23.3 0.95;
16 43.4 24.7 0.97;
17 45.7 25.1 0.96;
16 44.3 24.2 0.97;
16 43.6 25.3 0.97;
18 40.7 23.1 0.96;
17 40.2 22.8 0.95;
16 41.8 24.2 0.96;
18 42.9 24.7 0.96;
23 37.5 21.3 0.94;
22 36.8 21.2 0.93;
25 37.5 21.8 0.92;
24 38.3 35.4 0.93;
23 35.4 20.8 0.93;
23 37.3 21.3 0.92;
22 36.2 20.9 0.92;
24 36.8 21.3 0.88;
22 39.7 22.8 0.87;
22 37.2 21.4 0.86;
23 38.9 22.5 0.92;
23 39.2 23.1 0.92;
23 38.3 21.9 0.93;
24 36.4 21.2 0.93;
26 25.6 17.8 0.83;
28 27.4 18.7 0.85;
27 27.8 18.8 0.84;
28 32.3 20.2 0.82;
28 34.6 20.9 0.83;
29 29.2 19.2 0.80;
27 26.7 18.1 0.83;
28 27.9 18.9 0.82;
32 23.4 16.9 0.71;
35 22.7 16.7 0.73;
37 24.5 17.6 0.76;
33 20.8 15.8 0.78;
36 22.6 16.8 0.77;
34 21.4 16.4 0.77;
35 20.9 15.9 0.76];
y=sim(net,p_test');
请问怎么总是出现“Error using network/sim (line 130)
Input 1 size does not match net.inputs{1}.size.
%构建训练样本中的输入向量p
p1=[13 44.8 24.7 0.97;
14 42.3 24.4 0.95;
15 42.5 24.6 0.97;
17 45.6 25.8 0.96];
p2=[21 37.6 21.3 0.92;
25 37.6 21.7 0.90;
23 35.4 20.8 0.94;
24 37.2 21.6 0.87];
p3=[26 32.3 20.2 0.84;
27 27.6 18.7 0.81;
26 33.7 20.8 0.83;
29 32.3 20.3 0.83];
p4=[30 21.5 16.4 0.77;
32 20.6 15.8 0.76;
36 22.6 16.8 0.77;
35 21.3 16.2 0.75];
p=[p1 p2 p3 p4];
%构建训练样本中的目标向量t
t1=[1 1;1 1;1 1;1 1];
t2=[1 0;1 0;1 0;1 0];
t3=[0 1;0 1;0 1;0 1];
t4=[0 0;0 0;0 0;0 0];
t=[t1 t2 t3 t4];
p=p';t=t';
net=newff(p,t,3,{'tansig','logsig'},'traingd');
net.trainParam.epochs=500;
net.trainParam.goal=0.01;
net=train(net,p,t);
%测试样本进行分类
p_test=[13 44.7 24.3 0.95;
12 41.2 23.7 0.96;
13 42.5 24.6 0.97;
16 40.5 23.3 0.95;
16 43.4 24.7 0.97;
17 45.7 25.1 0.96;
16 44.3 24.2 0.97;
16 43.6 25.3 0.97;
18 40.7 23.1 0.96;
17 40.2 22.8 0.95;
16 41.8 24.2 0.96;
18 42.9 24.7 0.96;
23 37.5 21.3 0.94;
22 36.8 21.2 0.93;
25 37.5 21.8 0.92;
24 38.3 35.4 0.93;
23 35.4 20.8 0.93;
23 37.3 21.3 0.92;
22 36.2 20.9 0.92;
24 36.8 21.3 0.88;
22 39.7 22.8 0.87;
22 37.2 21.4 0.86;
23 38.9 22.5 0.92;
23 39.2 23.1 0.92;
23 38.3 21.9 0.93;
24 36.4 21.2 0.93;
26 25.6 17.8 0.83;
28 27.4 18.7 0.85;
27 27.8 18.8 0.84;
28 32.3 20.2 0.82;
28 34.6 20.9 0.83;
29 29.2 19.2 0.80;
27 26.7 18.1 0.83;
28 27.9 18.9 0.82;
32 23.4 16.9 0.71;
35 22.7 16.7 0.73;
37 24.5 17.6 0.76;
33 20.8 15.8 0.78;
36 22.6 16.8 0.77;
34 21.4 16.4 0.77;
35 20.9 15.9 0.76];
y=sim(net,p_test');
请问怎么总是出现“Error using network/sim (line 130)
Input 1 size does not match net.inputs{1}.size.
你的数据矩阵的组织出问题了,你设置的网络输入神经元和你每次输入的数据这块有矛盾,可以尝试先再减少数据量去做,这里有可能用到转职,然后再加大数据量,这样去组织数据