聚类算法 有监督类型的机器学习算法 一种思路:在构建大样本数据集的基础下, 然后利用特征工程+LR/xgboost/gbdt等进行, 或者直接采用用深度学习模型 另外一种思路:构建小样本数据集,利用active-learning
类似的一些新的研究也有很多,比如 Self-Ensembling Graph Convolutional Networks 、主动学习(Active-Learning)以及 A cluster-then-label
To reduce the dependency on labels, various active-learning strategies have been proposed, typically Experiments show that our method comprehensively outperforms a wide range of active-learning methods