Perceptron : Recognition Pattern ABC
4 views (last 30 days)
Show older comments
Hi, all
I have six patterns as shown below
A1 = [ -1 -1 1 1 -1 -1 -1;
-1 -1 -1 1 -1 -1 -1;
-1 -1 -1 1 -1 -1 -1;
-1 -1 1 -1 1 -1 -1;
-1 -1 1 -1 1 -1 -1;
-1 1 1 1 1 1 -1;
-1 1 -1 -1 -1 1 -1;
-1 1 -1 -1 -1 1 -1;
1 1 1 -1 1 1 1];
B1 = [ 1 1 1 1 1 1 1;
1 -1 -1 -1 -1 -1 1;
1 -1 -1 -1 -1 -1 1;
1 -1 -1 -1 -1 -1 1;
1 1 1 1 1 1 1;
1 -1 -1 -1 -1 -1 1;
1 -1 -1 -1 -1 -1 1;
1 -1 -1 -1 -1 -1 1;
1 1 1 1 1 1 1];
C1 = [ -1 -1 1 1 1 1 1 ;
-1 1 -1 -1 -1 -1 1;
1 -1 -1 -1 -1 -1 -1;
1 -1 -1 -1 -1 -1 -1;
1 -1 -1 -1 -1 -1 -1;
1 -1 -1 -1 -1 -1 -1;
1 -1 -1 -1 -1 -1 -1;
-1 1 -1 -1 -1 -1 1;
-1 -1 1 1 1 1 -1];
A2 = [ -1 -1 -1 1 -1 -1 -1;
-1 -1 -1 1 -1 -1 -1;
-1 -1 -1 1 -1 -1 -1;
-1 -1 1 -1 1 -1 -1;
-1 -1 1 -1 1 -1 -1;
-1 1 -1 -1 -1 1 -1;
-1 1 1 1 1 1 -1;
-1 1 -1 -1 -1 1 -1;
-1 1 -1 -1 -1 1 -1];
B2 = [ 1 1 1 1 1 1 -1;
1 -1 -1 -1 -1 -1 1;
1 -1 -1 -1 -1 -1 1;
1 -1 -1 -1 -1 -1 1;
1 1 1 1 1 1 -1;
1 -1 -1 -1 -1 -1 1;
1 -1 -1 -1 -1 -1 1;
1 -1 -1 -1 -1 -1 1;
1 1 1 1 1 1 -1];
C2 = [ -1 -1 1 1 1 -1 -1;
-1 1 -1 -1 -1 1 -1;
1 -1 -1 -1 -1 -1 1;
1 -1 -1 -1 -1 -1 -1;
1 -1 -1 -1 -1 -1 -1;
1 -1 -1 -1 -1 -1 -1;
1 -1 -1 -1 -1 -1 1;
-1 1 -1 -1 -1 1 -1;
-1 -1 1 1 1 -1 -1];
I have to recognize these patterns with artificial neural network.
I am new in Matlab. Please help!
I need to divide this data into 2 groups.
The first group A1, B1, C1 as training data. The second group A2, B2, C2 used to validate/test the network.
Example : if I select A1 then the output must display 'A', if I select B1 then the output must display 'B', if I select A2 then the output must display 'A'.
. . # # . . .
. . . # . . .
. . . # . . .
. . # . # . .
. . # . # . . => This pattern should be recognized as A
. # # # # # .
. # . . . # .
. # . . . # .
# # # . # # #
In result program, we must explain epochs from start to finish , which Learning Rate = 1 And Threshold Value = 0.5
How do I do that?
Thanks in advance!
Network type is perceptron
0 Comments
Accepted Answer
Greg Heath
on 13 Nov 2013
Start with MATLAB examples
help fitnet % regression/curve-fitting
help patternnet % classification/pattern-recognition
help nndata
Then submit failed code with comments and/or error messages
Hope this helps.
Greg
0 Comments
More Answers (1)
Greg Heath
on 13 Nov 2013
9X7 input matrices have to be columnized using the (:) operator into 63 dimensional vectors. Outputs should be columns of the 3-dimensional unit matrix. In order to avoid overtraining an overfit net, a tremendous amount of variable reduction should be applied to the inputs.
See Also
Products
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!