thomas's profile - activity

I have the following code:

CvSVMParams params;
params.svm_type = CvSVM::EPS_SVR;
params.kernel_type = CvSVM::RBF;
params.term_crit = TermCriteria(CV_TERMCRIT_ITER, (int)1e7, 1e-6);

CvSVM svm;
svm.train_auto(_data, _resp, _var_idx, _train_idx, params);

Here _data and _resp are Mats holding the feature vectors and responses, _var_idx containing the active features and _train_idx the active samples. For the parameter grids the default values are used. Unfortunately, the code produces the following error:

OpenCV Error: Assertion failed (sv_count != 0) in do_train, file /home/.../opencv-2.4.9/modules/ml/src/svm.cpp, line 1346

When I run a regression one the same data with parameters selected by hand it works fine. And when I switch to a classification problem (and change the corresponding parameters and SVM type) it also works. In that case for single training as well as for auto training. Also, when switching to CvSVM::NU_SVR the auto training works fine.

Another thing that bothers me is that I also have to provide the parameter p (EPS_EVR) or nu (NU_SVR) when I want to do auto training. The documentation says that those are also estimated using their corresponding default grids. Why is that so?

Thanks in advance.

EDIT:

I've made a small example that suffers from this problem. Just in case anyone want to try it out and reproduce the problem:

#include <iostream>
#include <opencv2/core/core.hpp>
#include <opencv2/ml/ml.hpp>

using namespace cv;
using namespace std;

int main() {

    Mat X(1000, 2, CV_32FC1);
    Mat Y(1000, 1, CV_32FC1);

    randu(X, -2, 2);

    for(int i = 0; i < 1000; i++){
        Y.at<float>(i,0) = pow(X.at<float>(i,0),2) + pow(X.at<float>(i,1),2) - 1;
    }

    CvSVMParams params;
    params.svm_type = CvSVM::EPS_SVR;
    params.kernel_type = CvSVM::RBF;
    params.term_crit = TermCriteria(CV_TERMCRIT_ITER, (int)1e7, 1e-6);
    params.p = 0.1;

    CvSVM svm;
    svm.train_auto(X, Y, Mat::ones(1,2, CV_8U), Mat::ones(1,1000, CV_8U), params);

    return 0;
}

This may seem like a duplicate question to this, but the difference is that there I was asking whether the output range is [-1,1] or [0,1]. I have accepted that the range is [0,1] if the the activation function is the sigmoid with alpha != 0 and beta != 0 (as stated in the documentation). Anyway, it seems to me that the output range is more like [0-eps, 1+eps].

My question is: Why is there a small epsilon and how can I turn this off?

One thing I could think of is that the output neurons aren't sigmoid units but linear units. Although it is explicitly stated that all neurons have the same activation function, this could explain this behavior.

Here is a small example that shows what I mean:

#include <iostream>
#include <opencv2/core/core.hpp>
#include <opencv2/ml/ml.hpp>

using namespace cv;
using namespace std;

int main() {

    int POS = 1, NEG = 0;

    int SAMPLES = 100;
    float SPLIT = 0.8;

    float C_X = 0.5;
    float C_Y = 0.5;
    float R = 0.3;

    Mat X(SAMPLES, 2, CV_32FC1);
    Mat Y(SAMPLES, 1, CV_32FC1);

    randu(X, 0, 1);

    for(int i = 0; i < SAMPLES; i++){
        Y.at<float>(i,0) = pow((X.at<float>(i,0) - C_X),2) + pow((X.at<float>(i,1) - C_Y),2) < pow(R,2) ? POS : NEG;
    }

    Mat X_train = X(Range(0, (int)(SAMPLES*SPLIT)), Range::all());
    Mat Y_train = Y(Range(0, (int)(SAMPLES*SPLIT)), Range::all());

    Mat X_test = X(Range((int)(SAMPLES*SPLIT), SAMPLES), Range::all());
    Mat Y_test = Y(Range((int)(SAMPLES*SPLIT), SAMPLES), Range::all());

    CvANN_MLP_TrainParams params(
                 cvTermCriteria(CV_TERMCRIT_ITER+CV_TERMCRIT_EPS, 1000, 0.000001),
                 CvANN_MLP_TrainParams::BACKPROP,
                 0.1,
                 0.1);

    Mat layers = (Mat_<int>(3,1) << 2, 4, 1);

    CvANN_MLP net(layers, CvANN_MLP::SIGMOID_SYM, 1, 1);
    net.train(X_train, Y_train, Mat(), Mat(), params);

    Mat predictions(Y_test.size(), CV_32F); 
    net.predict(X_test, predictions);

    cout << predictions << endl;

    Mat error = predictions-Y_test;
    multiply(error, error, error);

    float mse = sum(error)[0]/error.rows;

    cout << "MSE: " << mse << endl;

    return 0;
    }

For me this produces the following output:

[0.9940818;
0.087859474;
0.072328083;
0.032660298;
-0.0090373717;
0.056480117;
0.13302;
-0.025581671;
0.32763073;
1.0263158;
0.29676101;
0.056798562;
0.070351392;
1.0213233;
0.006240299;
0.96525788;
0.071746305;
1.0048869;
-0.015669812;
0.0023532249]
MSE: 0.0326775

As you can see, there are values just below 0 and above 1.