方法对比

	半监督梯度提升	Boosting
领域	机器学习	机器学习
方法族	Machine learning	Machine learning
起源年份	2006–2010s	1990–1997
提出者	Chapelle, Scholkopf & Zien (eds.); applied to GBM variants in subsequent literature	Schapire, R. E.; Freund, Y.
类型	Semi-supervised ensemble (self-training + gradient boosted trees)	Sequential ensemble (iterative reweighting)
开创性文献	Yarowsky, D. (1995). Unsupervised word sense disambiguation rivaling supervised methods. Proceedings of ACL 1995, 189–196. (Foundational self-training framework underlying pseudo-label approaches.)	Freund, Y. & Schapire, R. E. (1997). A decision-theoretic generalization of on-line learning and an application to boosting. Journal of Computer and System Sciences, 55(1), 119–139.
别名	pseudo-label gradient boosting, self-training GBM, semi-supervised GBT, label-propagation boosting	AdaBoost, gradient boosting, iterative reweighting ensemble, sequential ensemble
相关	6	6
摘要	Semi-supervised gradient boosting combines gradient boosted trees with self-training or pseudo-labeling to exploit large pools of unlabeled data alongside a small labeled set. An initial GBM fit on labeled data assigns confident predictions to unlabeled examples; those pseudo-labeled points are folded back into training and the model is re-boosted, iterating until convergence. This allows practitioners to harness cheap unlabeled data when labels are scarce or expensive.	Boosting is a sequential ensemble technique that converts many simple, barely-better-than-chance learners into a single highly accurate model by repeatedly focusing training on the examples that previous learners got wrong, then combining all learners with weights proportional to their individual accuracy.