Does prediction for a calculated svm model
[predicted_label, accuracy, decision_values] = svmpredict(testing_label_vector, testing_instance_matrix, model) [predicted_label, accuracy, decision_values] = svmpredict(testing_label_vector, testing_instance_matrix, model, 'libsvm_options')
SVM model structure from svmtrain.
-b probability_estimates: whether to predict probability estimates, 0 or 1 (default 0); one-class SVM not supported yet
SVM prediction output vector
a vector with accuracy, mean squared error, squared correlation coefficient.
If selected, probability estimate vector