0

I am trying to use y_scores=OneVsRestClassifier(svm.SVC()).predict() on datasets like iris and titanic .The trouble is that I am getting y_scores as continous values.like for iris dataset I am getting :

[[ -3.70047231  -0.74209097   2.29720159]
 [ -1.93190155   0.69106231  -2.24974856]
.....

I am using the OneVsRestClassifier for other classifier models like knn,randomforest,naive bayes and they are giving appropriate results in the form of 

[[ 0  1   0]
 [ 1   0  1]...

etc on the iris dataset .Please help.

alexander
  • 21
  • 7

1 Answers1

1

Well this is simply not true.

>>> from sklearn.multiclass import OneVsRestClassifier
>>> from sklearn.svm import SVC
>>> from sklearn.datasets import load_iris
>>> iris = load_iris()
>>> clf = OneVsRestClassifier(SVC())
>>> clf.fit(iris['data'], iris['target'])
OneVsRestClassifier(estimator=SVC(C=1.0, cache_size=200, class_weight=None, coef0=0.0, degree=3, gamma=0.0,
  kernel='rbf', max_iter=-1, probability=False, random_state=None,
  shrinking=True, tol=0.001, verbose=False),
          n_jobs=1)
>>> print clf.predict(iris['data'])
[0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
 0 0 0 0 0 0 0 0 0 0 0 0 0 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 2 1 1 1 1 1 2 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2
 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2
 2 2]

maybe you called decision_function instead (which would match your output dimension, as predict is supposed to return a vector, not a matrix). Then, SVM returns signed distances to each hyperplane, which is its decision function from mathematical perspective.

lejlot
  • 64,777
  • 8
  • 131
  • 164