Why is neural network not predicting?

Question

I am trying to train a neural network using backpropagation algo. in OpenCV 2.3. However it is not predicting correctly....not even on training dataset. Could anybody please help me find whats wrong here?

training_feature_matrix - Nx69 matrix of float values

training_age_matrix - Nx4 matrix of float values

test_feature_matrix - Mx69 matrix of float values

test_age_matrix - Mx4 matrix of float values

the feature matrices (mentioned above) are like: [0.123435, 0.4542665, 0.587545, ...68-such values + last value '1.0 or 2.0' depending upon its male/female)

the age-matrices (mentioned above) are like: [1, 0, 0 ,0; 1, 0, 0, 0; 0, 1, 0, 0; ...] here 1s show the class of age (baby, child, adult, old) the corresponding row of feature matrix belongs to.

here is the code: I call 'mlp' function using above matrices as parameters)

cv::Mat mlp(cv::Mat& training_feature_matrix, cv::Mat& training_age_matrix, cv::Mat& test_feature_matrix, cv::Mat& test_age_matrix)
{
cv::Mat layers = cv::Mat(3, 1, CV_32SC1);
layers.row(0)  = cv::Scalar(69);
layers.row(1)  = cv::Scalar(36);
layers.row(2)  = cv::Scalar(4);    //   cout<<layers<<"\n";

CvANN_MLP ann;
CvANN_MLP_TrainParams params;
CvTermCriteria criteria;
criteria.max_iter = 10000;
criteria.epsilon  = 0.001;
criteria.type     = CV_TERMCRIT_ITER + CV_TERMCRIT_EPS;
params.train_method = CvANN_MLP_TrainParams::BACKPROP;
params.bp_dw_scale  = 0.1;
params.bp_moment_scale = 0.1;
params.term_crit  = criteria;

ann.create(layers, CvANN_MLP::SIGMOID_SYM);
ann.train(training_feature_matrix, training_age_matrix, cv::Mat(), cv::Mat(), params);

cv::Mat predicted(test_age_matrix.rows, 4, CV_32SC1);
for(int i = 0; i < test_feature_matrix.rows; i++)
{
  cv::Mat response(1, 4, CV_32F);
  cv::Mat sample = test_feature_matrix.row(i);
  ann.predict(sample, response);
  for (int g = 0; g < 4; g++)
  {
    predicted.at<int>(i,g) = response.at<float>(0,g);
  } 
}
   cout << "\n";
   cout << ann.get_weights(0) << "\n";
   cout << ann.get_layer_sizes() << "\n";
   cout << ann.get_layer_count() << "\n\n";

return predicted;
}

EDIT Also, the ann.get_weights(0) & ann.get_layer_sizes() are returning garbage values but ann.get_layer_count() is returning correct value 3.

Thanks :)

ann.get_weights(0) & ann.get_layer_sizes() return pointers, so will look like "rubbish" if you print them like you have. The rest of your code seems to be ok, are you sure your data is good? What exactly does "it is not predicting correctly" mean? — Bull, May 17 '13 at 11:08
@user2151446 how do i extract values from those pointers? not predicting correctly means....the output prediction matrix that I'm getting is a matrix of float values...there are positive, negative values and also values greater than 1....its not making sense... the input data is perfectly fine...each in the form of float-matrix as OpenCV demands... — learner, May 17 '13 at 11:17
@user2151446 I am able to extract values out from ann.get_weights() but not bale to extract ann.get_layer_sizes() value. What should I do? — learner, May 17 '13 at 11:30
Is there a particular reason for using 2.3 version? Could you give a try with the most recent 2.4.6? — gpicchiarelli, Jul 09 '13 at 12:28
Why do you convert roughly your result into int? I suppose your result is a matrix full of zeros... — Bentoy13, Jul 19 '13 at 12:35
@learner: Is there a specific reason for giving layers 3 rows and 1 column ? I thought that this matrix had only one row. — Fabien R, Aug 03 '13 at 13:39

score 2 · Answer 1 · edited May 23 '17 at 12:22

2

It has been long since that question asked but I will share the answer. I had a similar problem with sigmoid's output values. It is resolved now. You can check my issue here :

OpenCV Neural Network Sigmoid Output

To summarize the error, it is occuring because of the default parameters of mlp's create function. Use like this : ann.create(layers, CvANN_MLP::SIGMOID_SYM, 1, 1).

edited May 23 '17 at 12:22

Community

1
1

answered Feb 28 '14 at 07:45

yutasrobot

2,356
1
17
24

score 1 · Answer 2 · answered Aug 03 '13 at 22:46

1

Back propagation does not always converge. It is quite likely to blow up and produce nonsense. This is likely if the epsilon or momentum_scale values are too large. Your momentum looks to be at the top end of what might work and I would try reducing it.

answered Aug 03 '13 at 22:46

David Elliman

1,379
8
15

Why is neural network not predicting?

2 Answers2