使用无监督学习算法对标签数据。
半监督学习算法同时考虑有标记和无标记数据，能显著提升学习效果。
Semi-supervised learning algorithms, which consider both labeled and unlabeled data, can improve learning effectiveness significantly.
然而，传统的监督学习算法需要标记大量的训练样本来建立满意的分类器。
However, traditional supervised learning techniques typically require a large number of labeled examples to learn an accurate classifier.
应用推荐