ScholarGate
Asystent

Porównaj metody

Przeglądaj wybrane metody obok siebie; wiersze, które się różnią, są wyróżnione.

Uczenie metryczne wspomagane częściowym nadzorem×Uczenie ze wsparciem częściowym×
DziedzinaUczenie maszynoweUczenie maszynowe
RodzinaMachine learningMachine learning
Rok powstania2007–20081970s–2006 (formalized)
TwórcaYeung, D.-Y. & Chang, H.; Davis, J. V. & Dhillon, I. S.Vapnik, V. N. and others (community of researchers, 1970s–2000s)
TypHybrid supervised/unsupervised distance learningLearning paradigm
Źródło pierwotneYeung, D.-Y., & Chang, H. (2007). A kernel approach for semi-supervised metric learning. IEEE Transactions on Neural Networks, 18(1), 141–149. DOI ↗Chapelle, O., Scholkopf, B., & Zien, A. (Eds.) (2006). Semi-Supervised Learning. MIT Press. ISBN: 978-0-262-03358-9
Inne nazwySSML, semi-supervised distance learning, constrained metric learning, weakly supervised metric learningSSL, semi-supervised machine learning, transductive learning, label-efficient learning
Pokrewne55
PodsumowanieSemi-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.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.
ScholarGateZbiór danych
  1. v1
  2. 2 Źródła
  3. PUBLISHED
  1. v1
  2. 2 Źródła
  3. PUBLISHED

Przejdź do wyszukiwania Pobierz slajdy

ScholarGatePorównaj metody: Semi-supervised Metric Learning · Semi-supervised Learning. Pobrano 2026-06-17 z https://scholargate.app/pl/compare