方法对比

	Ensemble Online Learning	Boosting	在线学习
领域	机器学习	机器学习	机器学习
方法族	Machine learning	Machine learning	Machine learning
起源年份	2001	1990–1997	1958–2000s
提出者	Oza, N. C. & Russell, S.	Schapire, R. E.; Freund, Y.	Rosenblatt, F.; Littlestone, N.; Shalev-Shwartz, S. (key contributors)
类型	Ensemble (online / incremental)	Sequential ensemble (iterative reweighting)	Learning paradigm (sequential model update)
开创性文献	Oza, N. C., & Russell, S. (2001). Online bagging and boosting. In Proceedings of the Eighth International Workshop on Artificial Intelligence and Statistics (AISTATS 2001), pp. 229–236.	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.	Shalev-Shwartz, S. (2011). Online Learning and Online Convex Optimization. Foundations and Trends in Machine Learning, 4(2), 107–194.
别名	online ensemble methods, streaming ensemble learning, incremental ensemble learning, adaptive ensemble learning	AdaBoost, gradient boosting, iterative reweighting ensemble, sequential ensemble	incremental learning, sequential learning, streaming learning, online machine learning
相关	6	6	6
摘要	Ensemble Online Learning combines multiple base learners that are trained incrementally on a stream of data, updating each model one observation at a time. By aggregating the predictions of diverse online learners, the ensemble achieves accuracy and robustness that surpass any single incremental model, while adapting continuously to changing data distributions.	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.	Online learning is a machine learning paradigm in which a model is updated incrementally as each new data point arrives, rather than being trained once on a fixed dataset. It is essential when data streams continuously, storage is limited, or the underlying distribution shifts over time. Theoretical performance is measured by cumulative regret relative to the best fixed predictor in hindsight.