Methoden vergleichen

	Bayesian Bagging	Bayesian Boosting	Boosting
Fachgebiet	Maschinelles Lernen	Maschinelles Lernen	Maschinelles Lernen
Familie	Machine learning	Machine learning	Machine learning
Entstehungsjahr	2001	1999–2010	1990–1997
Urheber	Clyde, M. & Lee, H. (building on Rubin's Bayesian bootstrap, 1981)	Ridgeway, G.; Chipman, H. A. et al.	Schapire, R. E.; Freund, Y.
Typ	Ensemble (Bayesian bootstrap aggregation)	Probabilistic ensemble (Bayesian interpretation of boosting)	Sequential ensemble (iterative reweighting)
Wegweisende Quelle	Clyde, M. & Lee, H. (2001). Bagging and the Bayesian bootstrap. In T. Richardson & T. Jaakkola (Eds.), Proceedings of the Eighth International Workshop on Artificial Intelligence and Statistics (AISTATS 2001).	Ridgeway, G. (1999). The state of boosting. Computing Science and Statistics, 31, 172–181.	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.
Aliasnamen	Bayesian bootstrap aggregation, BB-ensemble, Bayesian model averaging via bootstrap, Bayesian bagged ensemble	Bayesian ensemble boosting, probabilistic boosting, Bayesian additive model, Bayesian boosted ensemble	AdaBoost, gradient boosting, iterative reweighting ensemble, sequential ensemble
Verwandt	6	5	6
Zusammenfassung	Bayesian Bagging replaces the classical bootstrap with the Bayesian bootstrap — drawing Dirichlet-distributed weights over training observations rather than sampling with replacement — and trains an ensemble of base learners under those weights. The result is a principled ensemble that approximates a Bayesian posterior over predictions, yielding calibrated uncertainty estimates alongside strong predictive accuracy.	Bayesian boosting integrates probabilistic Bayesian inference with boosting ensemble techniques, combining multiple weak learners while maintaining full uncertainty quantification over predictions. Unlike standard gradient boosting that produces a single point estimate, Bayesian boosting yields a posterior distribution over the ensemble output, enabling calibrated confidence intervals alongside predictions.	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.