Σύγκριση μεθόδων

	Ημι-εποπτευόμενη Ενίσχυση (Semi-supervised Boosting)	Ενίσχυση Κλίσης (Gradient Boosting)	Ημι-επιβλεπόμενη Μάθηση
Πεδίο	Μηχανική Μάθηση	Μηχανική Μάθηση	Μηχανική Μάθηση
Οικογένεια	Machine learning	Machine learning	Machine learning
Έτος προέλευσης	1999–2009	2001	1970s–2006 (formalized)
Δημιουργός	Mallapragada, P. K.; Bennett, K. P.; and others	Friedman, J. H.	Vapnik, V. N. and others (community of researchers, 1970s–2000s)
Τύπος	Semi-supervised ensemble method	Ensemble (sequential boosting of decision trees)	Learning paradigm
Θεμελιώδης πηγή	Mallapragada, P. K., Jin, R., Jain, A. K., & Liu, Y. (2009). SemiBoost: Boosting for Semi-supervised Learning. IEEE Transactions on Pattern Analysis and Machine Intelligence, 31(11), 2000–2014.	Friedman, J. H. (2001). Greedy Function Approximation: A Gradient Boosting Machine. Annals of Statistics, 29(5), 1189–1232.	Chapelle, O., Scholkopf, B., & Zien, A. (Eds.) (2006). Semi-Supervised Learning. MIT Press. ISBN: 978-0-262-03358-9
Εναλλακτικές ονομασίες	SemiBoost, SSL boosting, boosting with unlabeled data, semi-supervised ensemble boosting	Gradient Boosting (GBM), GBM, gradient boosted trees, gradient boosting machine	SSL, semi-supervised machine learning, transductive learning, label-efficient learning
Συναφείς	5	5	5
Σύνοψη	Semi-supervised Boosting is an ensemble learning paradigm that extends classical boosting algorithms — such as AdaBoost — to exploit both labeled and unlabeled data. By propagating label information through a similarity structure over unlabeled instances, it trains stronger classifiers than supervised boosting alone when labeled data are scarce.	Gradient Boosting is an ensemble learning method, formalised by Jerome H. Friedman in 2001, that combines a sequence of weak learners — typically shallow decision trees — so that each new tree is fitted to minimise the residual errors of the trees before it. It is the core algorithm behind popular implementations such as XGBoost, LightGBM and CatBoost.	Semi-supervised learning (SSL) is a machine learning paradigm that trains models using a small set of labeled examples together with a much larger pool of unlabeled data. By leveraging the structure inherent in unlabeled data, SSL achieves accuracy closer to fully supervised models while requiring far fewer costly manual labels — making it practical when labeling is expensive, slow, or resource-constrained.