Krahasoni metodat

	Ansambli i Stacking-ut Bajesian	Boosting	Procesi Gaussian
Fusha	Mësimi i makinës	Mësimi i makinës	Mësimi i makinës
Familja	Machine learning	Machine learning	Machine learning
Viti i origjinës	2018	1990–1997	2006 (book); roots in Kriging, 1951)
Krijuesi	Yao, Y.; Vehtari, A.; Simpson, D.; Gelman, A.	Schapire, R. E.; Freund, Y.	Rasmussen, C. E. & Williams, C. K. I.
Lloji	Bayesian ensemble combination	Sequential ensemble (iterative reweighting)	Probabilistic non-parametric model
Burimi themelues	Yao, Y., Vehtari, A., Simpson, D., & Gelman, A. (2018). Using stacking to average Bayesian predictive distributions. Bayesian Analysis, 13(3), 917–1007.	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.	Rasmussen, C. E., & Williams, C. K. I. (2006). Gaussian Processes for Machine Learning. MIT Press. ISBN: 978-0-262-18253-9
Emërtime të tjera	Bayesian stacking, Bayesian model stacking, stacking with Bayesian weights, predictive distribution stacking	AdaBoost, gradient boosting, iterative reweighting ensemble, sequential ensemble	GP, Gaussian Process Regression, GPR, Kriging
Të lidhura	6	6	3
Përmbledhja	Bayesian stacking combines the predictive distributions of several base models by finding non-negative weights that maximise the leave-one-out log predictive score of the mixture. Formalised by Yao, Vehtari, Simpson, and Gelman (2018), it yields a single calibrated predictive distribution that is provably at least as good as any single constituent model under cross-validation.	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.	A Gaussian Process (GP) is a non-parametric, fully probabilistic machine learning model that places a prior distribution directly over functions. Rather than predicting a single value, it returns a predictive mean and a calibrated uncertainty estimate at every test point, making it especially valuable for regression on small to medium datasets and for Bayesian optimization tasks.