Pattern Recognition Neural Network

3 次查看（过去 30 天）
massimiliano de martino
Hello to everyboody, I am tring to create a NN able to recognize a pattern. In parcticular input pattern is composed by 3 families and each family has been characterized by ten characters (attributes).
Family 1 Family 2 Family 3
0,956617957691530 0,490089874158284 0,326312311629368
0,544128360058316 1,05676774443777 0,440772213112043
0,832147860753549 0,494254930533971 0,171741757138505
0,240120429747655 0,637745305127687 0,547561629714484
0,447160087051316 1,23865447004323 0,426986449178013
0,269681946119199 0,418739260030671 0,393573860121322
0,717662542640876 0,963189757946470 0,271209060303500
0,444603891423024 0,826265869800830 0,313549793727008
0,309088072079284 0,483085778457597 0,685152011149895
0,612627689843110 0,630930803284252 0,653057923052316
Ouput is composed by a matrix where each column indicates a category (each category has been composed by a vector of 1 or zero).
Pattern1 Pattern2 Pattern3
0 1 0
0 0 0
1 0 1
1 1 1
1 0 1
1 1 0
0 0 1
0 1 1
1 1 0
0 0 0
Here I post algorithm created :
hiddenLayerSize = 20;
net = patternnet(hiddenLayerSize);
net.divideParam.trainRatio = 70/100;
net.divideParam.valRatio = 15/100;
net.divideParam.testRatio = 15/100;
[net] = train(net,input_NN,output_NN);
outputs_Calculate = net(Input_Test)
In the last row i've tested NN to recognize a Pattern on which NN has been trained (parcticularly Family 1),but results is not good :
Input Test Output Test
0,956617957691530 2,02003965807583e-10
0,544128360058316 0
0,832147860753549 0,000250000050416781
0,240120429747655 0,00100000000000000
0,447160087051316 0,000249999962817318
0,269681946119199 0,000249999745213518
0,717662542640876 2,26148556268713e-10
0,444603891423024 1,78739533980832e-10
0,309088072079284 0,000249999634660327
0,612627689843110 0
NN is not able to recognize the pattern. Could someone suggests me how to solve this problem ???

请先登录，再回答此问题。

回答（0 个）

类别

在 Help Center 和 File Exchange 中查找有关 Pattern Recognition and Classification 的更多信息

标签

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by