For optimal performance, use C-ordered numpy.ndarray (dense) or ... get number of support vectors for each class >>> clf.n_support_ array([1, 1]...).
確定! 回上一頁