神经网络PNN问题 - MATLAB爱好者论坛-LabFans.com

wsy1351 · 2009-05-09, 17:05

%输入向量P和目标向量T
P=[1.033 0.197 0.185 0.306;
1.045 0.205 0.309 0.354;
1.041 0.194 0.455 0.406;
1.046 0.194 0.587 0.454;
1.061 0.193 0.719 0.504;
1.042 0.367 0.102 0.304;
1.050 0.386 0.249 0.354;
1.064 0.370 0.407 0.406;
1.081 0.361 0.698 0.506;
1.033 0.234 0.743 0.507;
1.062 0.235 0.598 0.454;
1.055 0.242 0.467 0.407;
1.052 0.241 0.318 0.355;
1.047 0.247 0.176 0.305;
1.041 0.250 0.032 0.252;
1.050 0.338 0.126 0.307;
1.060 0.377 0.267 0.355;
1.060 0.340 0.419 0.403;
1.068 0.331 0.561 0.454;
1.026 0.196 0.242 0.354;
1.033 0.181 0.370 0.407;
1.046 0.180 0.487 0.456;
1.053 0.178 0.627 0.508;
1.033 0.362 0.031 0.303;
1.043 0.370 0.171 0.354;
1.050 0.362 0.312 0.406;
1.061 0.367 0.436 0.455;
1.057 0.359 0.572 0.506;
1.068 0.241 0.712 0.517;
1.069 0.248 0.631 0.488;
1.056 0.245 0.503 0.437;
1.049 0.248 0.353 0.382;
1.053 0.249 0.204 0.327;
1.055 0.336 0.157 0.328;
1.057 0.336 0.305 0.379;
1.060 0.341 0.419 0.429;
1.067 0.333 0.600 0.484;
1.075 0.332 0.501 0.516]';
T=[1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
0 0 1;
0 0 1;
0 0 1;
0 0 1;
0 0 1;
0 0 1;
0 0 1;
0 0 1;
0 0 1;
0 1 0;
0 1 0;
0 1 0;
0 1 0;
0 1 0;
0 1 0;
0 1 0;
0 1 0;
0 1 0;
0 1 0]';
%创建4个PNN，SPREAD不同
for i=1:4
net=newpnn(P,T,i/8);
temp=sim(net,P)
yc=vec2ind(temp)
end
%测试样本
P_test=[1.020 0.190 0.052 0.255;
1.085 0.330 0.701 0.502;
1.028 0.184 0.122 0.306;
1.044 0.251 0.053 0.273]';
y_test=sim(net,P_test);
yc_test=vec2ind(y_test);
yc_test
谁能告诉我为什么会
Error in ==> PNN1 at 81
net=newpnn(P,T,i/8);

karl_wang · 2009-05-09, 20:51

%输入向量P和目标向量T
P=[1.033 0.197 0.185 0.306;
1.045 0.205 0.309 0.354;
1.041 0.194 0.455 0.406;
1.046 0.194 0.587 0.454;
1.061 0.193 0.719 0.504;
1.042 0.367 0.102 0.304;
1.050 0.386 0.249 0.354;
1.064 0.370 0.407 0.406;
1.081 0.361 0.698 0.506;
1.033 0.234 0.743 0.507;
1.062 0.235 0.598 0.454;
1.055 0.242 0.467 0.407;
1.052 0.241 0.318 0.355;
1.047 0.247 0.176 0.305;
1.041 0.250 0.032 0.252;
1.050 0.338 0.126 0.307;
1.060 0.377 0.267 0.355;
1.060 0.340 0.419 0.403;
1.068 0.331 0.561 0.454;
1.026 0.196 0.242 0.354;
1.033 0.181 0.370 0.407;
1.046 0.180 0.487 0.456;
1.053 0.178 0.627 0.508;
1.033 0.362 0.031 0.303;
1.043 0.370 0.171 0.354;
1.050 0.362 0.312 0.406;
1.061 0.367 0.436 0.455;
1.057 0.359 0.572 0.506;
1.068 0.241 0.712 0.517;
1.069 0.248 0.631 0.488;
1.056 0.245 0.503 0.437;
1.049 0.248 0.353 0.382;
1.053 0.249 0.204 0.327;
1.055 0.336 0.157 0.328;
1.057 0.336 0.305 0.379;
1.060 0.341 0.419 0.429;
1.067 0.333 0.600 0.484;
1.075 0.332 0.501 0.516]';
T=[1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
1 0 0;
0 0 1;
0 0 1;
0 0 1;
0 0 1;
0 0 1;
0 0 1;
0 0 1;
0 0 1;
0 0 1;
0 1 0;
0 1 0;
0 1 0;
0 1 0;
0 1 0;
0 1 0;
0 1 0;
0 1 0;
0 1 0]';
%创建4个PNN，SPREAD不同
for i=1:4
net=newpnn(P,T,i/8);
temp=sim(net,P);
yc=vec2ind(temp)
end
%测试样本
P_test=[1.020 0.190 0.052 0.255;
1.085 0.330 0.701 0.502;
1.028 0.184 0.122 0.306;
1.044 0.251 0.053 0.273]';
y_test=sim(net,P_test);
yc_test=vec2ind(y_test);
yc_test
你再运行下，可以出结果！

wsy1351 · 2009-05-10, 07:44

还是不行啊。。。它显示??? Error using ==> newpnn
Inputs and Targets have different numbers of columns.

Error in ==> PNN1 at 81
net=newpnn(P,T,i/8);
这和MATLAB版本有关系吗？我的是7.0的

xyxcz · 2009-05-10, 22:26

yc =

Columns 1 through 11

1 1 1 1 1 1 1 1 1 1 1

Columns 12 through 22

1 1 1 1 1 1 1 1 1 1 1

Columns 23 through 33

1 1 1 1 1 1 1 1 1 1 1

Columns 34 through 38

1 1 1 1 2

yc =

Columns 1 through 11

1 1 1 1 1 1 1 1 1 1 1

Columns 12 through 22

1 1 1 1 1 1 1 1 1 1 1

Columns 23 through 33

1 1 1 1 1 1 1 1 1 1 1

Columns 34 through 38

1 1 1 1 1

yc =

Columns 1 through 11

1 1 1 1 1 1 1 1 1 1 1

Columns 12 through 22

1 1 1 1 1 1 1 1 1 1 1

Columns 23 through 33

1 1 1 1 1 1 1 1 1 1 1

Columns 34 through 38

1 1 1 1 1

yc =

Columns 1 through 11

1 1 1 1 1 1 1 1 1 1 1

Columns 12 through 22

1 1 1 1 1 1 1 1 1 1 1

Columns 23 through 33

1 1 1 1 1 1 1 1 1 1 1

Columns 34 through 38

1 1 1 1 1

yc_test =

1 1 1 1

我也是7.0版本

karl_wang · 2009-05-11, 20:27

引用:

作者: wsy1351

还是不行啊。。。它显示??? Error using ==> newpnn
Inputs and Targets have different numbers of columns.

Error in ==> PNN1 at 81
net=newpnn(P,T,i/8);
这和MATLAB版本有关系吗？我的是7.0的

我的可以运行，没有问题！我的是7.1版本！

ashhell · 2009-06-08, 15:42

回复2楼

我用MATLAB7.6 即2008a，运行结果如下

>> yc_test

yc_test =

1 1 1 1

是不是不正确？

karl_wang · 2009-06-09, 11:40

没有问题。你可以查询下vec2ind的用法。

2009-05-09, 17:05	#1
wsy1351 初级会员注册日期: 2009-05-09 帖子: 2 声望力: 0	神经网络PNN问题 %输入向量P和目标向量T P=[1.033 0.197 0.185 0.306; 1.045 0.205 0.309 0.354; 1.041 0.194 0.455 0.406; 1.046 0.194 0.587 0.454; 1.061 0.193 0.719 0.504; 1.042 0.367 0.102 0.304; 1.050 0.386 0.249 0.354; 1.064 0.370 0.407 0.406; 1.081 0.361 0.698 0.506; 1.033 0.234 0.743 0.507; 1.062 0.235 0.598 0.454; 1.055 0.242 0.467 0.407; 1.052 0.241 0.318 0.355; 1.047 0.247 0.176 0.305; 1.041 0.250 0.032 0.252; 1.050 0.338 0.126 0.307; 1.060 0.377 0.267 0.355; 1.060 0.340 0.419 0.403; 1.068 0.331 0.561 0.454; 1.026 0.196 0.242 0.354; 1.033 0.181 0.370 0.407; 1.046 0.180 0.487 0.456; 1.053 0.178 0.627 0.508; 1.033 0.362 0.031 0.303; 1.043 0.370 0.171 0.354; 1.050 0.362 0.312 0.406; 1.061 0.367 0.436 0.455; 1.057 0.359 0.572 0.506; 1.068 0.241 0.712 0.517; 1.069 0.248 0.631 0.488; 1.056 0.245 0.503 0.437; 1.049 0.248 0.353 0.382; 1.053 0.249 0.204 0.327; 1.055 0.336 0.157 0.328; 1.057 0.336 0.305 0.379; 1.060 0.341 0.419 0.429; 1.067 0.333 0.600 0.484; 1.075 0.332 0.501 0.516]'; T=[1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 0 0 1; 0 0 1; 0 0 1; 0 0 1; 0 0 1; 0 0 1; 0 0 1; 0 0 1; 0 0 1; 0 1 0; 0 1 0; 0 1 0; 0 1 0; 0 1 0; 0 1 0; 0 1 0; 0 1 0; 0 1 0; 0 1 0]'; %创建4个PNN，SPREAD不同 for i=1:4 net=newpnn(P,T,i/8); temp=sim(net,P) yc=vec2ind(temp) end %测试样本 P_test=[1.020 0.190 0.052 0.255; 1.085 0.330 0.701 0.502; 1.028 0.184 0.122 0.306; 1.044 0.251 0.053 0.273]'; y_test=sim(net,P_test); yc_test=vec2ind(y_test); yc_test 谁能告诉我为什么会 Error in ==> PNN1 at 81 net=newpnn(P,T,i/8);

2009-05-09, 20:51	#2
karl_wang 版主注册日期: 2008-12-03 年龄: 41 帖子: 43 声望力: 17	回复: 神经网络PNN问题 %输入向量P和目标向量T P=[1.033 0.197 0.185 0.306; 1.045 0.205 0.309 0.354; 1.041 0.194 0.455 0.406; 1.046 0.194 0.587 0.454; 1.061 0.193 0.719 0.504; 1.042 0.367 0.102 0.304; 1.050 0.386 0.249 0.354; 1.064 0.370 0.407 0.406; 1.081 0.361 0.698 0.506; 1.033 0.234 0.743 0.507; 1.062 0.235 0.598 0.454; 1.055 0.242 0.467 0.407; 1.052 0.241 0.318 0.355; 1.047 0.247 0.176 0.305; 1.041 0.250 0.032 0.252; 1.050 0.338 0.126 0.307; 1.060 0.377 0.267 0.355; 1.060 0.340 0.419 0.403; 1.068 0.331 0.561 0.454; 1.026 0.196 0.242 0.354; 1.033 0.181 0.370 0.407; 1.046 0.180 0.487 0.456; 1.053 0.178 0.627 0.508; 1.033 0.362 0.031 0.303; 1.043 0.370 0.171 0.354; 1.050 0.362 0.312 0.406; 1.061 0.367 0.436 0.455; 1.057 0.359 0.572 0.506; 1.068 0.241 0.712 0.517; 1.069 0.248 0.631 0.488; 1.056 0.245 0.503 0.437; 1.049 0.248 0.353 0.382; 1.053 0.249 0.204 0.327; 1.055 0.336 0.157 0.328; 1.057 0.336 0.305 0.379; 1.060 0.341 0.419 0.429; 1.067 0.333 0.600 0.484; 1.075 0.332 0.501 0.516]'; T=[1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 1 0 0; 0 0 1; 0 0 1; 0 0 1; 0 0 1; 0 0 1; 0 0 1; 0 0 1; 0 0 1; 0 0 1; 0 1 0; 0 1 0; 0 1 0; 0 1 0; 0 1 0; 0 1 0; 0 1 0; 0 1 0; 0 1 0]'; %创建4个PNN，SPREAD不同 for i=1:4 net=newpnn(P,T,i/8); temp=sim(net,P); yc=vec2ind(temp) end %测试样本 P_test=[1.020 0.190 0.052 0.255; 1.085 0.330 0.701 0.502; 1.028 0.184 0.122 0.306; 1.044 0.251 0.053 0.273]'; y_test=sim(net,P_test); yc_test=vec2ind(y_test); yc_test 你再运行下，可以出结果！

2009-05-10, 07:44	#3
wsy1351 初级会员注册日期: 2009-05-09 帖子: 2 声望力: 0	回复: 神经网络PNN问题还是不行啊。。。它显示??? Error using ==> newpnn Inputs and Targets have different numbers of columns. Error in ==> PNN1 at 81 net=newpnn(P,T,i/8); 这和MATLAB版本有关系吗？我的是7.0的

2009-05-10, 22:26	#4
xyxcz 初级会员注册日期: 2009-03-21 年龄: 39 帖子: 9 声望力: 0	回复: 神经网络PNN问题 yc = Columns 1 through 11 1 1 1 1 1 1 1 1 1 1 1 Columns 12 through 22 1 1 1 1 1 1 1 1 1 1 1 Columns 23 through 33 1 1 1 1 1 1 1 1 1 1 1 Columns 34 through 38 1 1 1 1 2 yc = Columns 1 through 11 1 1 1 1 1 1 1 1 1 1 1 Columns 12 through 22 1 1 1 1 1 1 1 1 1 1 1 Columns 23 through 33 1 1 1 1 1 1 1 1 1 1 1 Columns 34 through 38 1 1 1 1 1 yc = Columns 1 through 11 1 1 1 1 1 1 1 1 1 1 1 Columns 12 through 22 1 1 1 1 1 1 1 1 1 1 1 Columns 23 through 33 1 1 1 1 1 1 1 1 1 1 1 Columns 34 through 38 1 1 1 1 1 yc = Columns 1 through 11 1 1 1 1 1 1 1 1 1 1 1 Columns 12 through 22 1 1 1 1 1 1 1 1 1 1 1 Columns 23 through 33 1 1 1 1 1 1 1 1 1 1 1 Columns 34 through 38 1 1 1 1 1 yc_test = 1 1 1 1 我也是7.0版本

2009-06-08, 15:42	#6
ashhell 初级会员注册日期: 2008-09-07 年龄: 40 帖子: 1 声望力: 0	回复: 神经网络PNN问题回复2楼我用MATLAB7.6 即2008a，运行结果如下 >> yc_test yc_test = 1 1 1 1 是不是不正确？

2009-06-09, 11:40	#7
karl_wang 版主注册日期: 2008-12-03 年龄: 41 帖子: 43 声望力: 17	回复: 神经网络PNN问题没有问题。你可以查询下vec2ind的用法。