Comparer des méthodes

	Apprentissage métrique robuste	Apprentissage métrique semi-supervisé
Domaine	Apprentissage automatique	Apprentissage automatique
Famille	Machine learning	Machine learning
Année d'origine	2009–2012	2007–2008
Auteur d'origine	Various (Weinberger, Saul, Schultz et al.; robust extensions by Shen, Cao and others, 2009–2012)	Yeung, D.-Y. & Chang, H.; Davis, J. V. & Dhillon, I. S.
Type	Supervised/semi-supervised distance metric learning with robustness to noise and outliers	Hybrid supervised/unsupervised distance learning
Source fondatrice	Shen, C., Kim, J., Wang, L., & van den Hengel, A. (2012). Positive Semidefinite Metric Learning Using Boosting-like Algorithms. Journal of Machine Learning Research, 13, 1007–1036.	Yeung, D.-Y., & Chang, H. (2007). A kernel approach for semi-supervised metric learning. IEEE Transactions on Neural Networks, 18(1), 141–149.
Alias	robust distance metric learning, noise-robust metric learning, outlier-robust similarity learning, robust DML	SSML, semi-supervised distance learning, constrained metric learning, weakly supervised metric learning
Apparentées	5	5
Résumé	Robust Metric Learning learns a Mahalanobis distance function from labeled or pairwise-constrained data while actively resisting the distortion caused by noisy labels, corrupted examples, or outliers. By replacing standard hinge or squared losses with robust alternatives and adding regularization, it produces a distance metric that generalises well even when the training set is imperfect — a common situation in real-world scientific and applied tasks.	Semi-supervised metric learning learns a task-adapted distance function by combining a small set of labeled pairwise constraints — must-link and cannot-link pairs — with the geometric structure of a much larger pool of unlabeled data. The result is a Mahalanobis-style or kernel-based distance that reflects both supervision and data topology, improving downstream tasks such as nearest-neighbor classification and clustering.