Sammenlign metoder

	Selv-overvåget K-nærmeste naboer	Overførselslæring
Fagområde	Maskinlæring	Maskinlæring
Familie	Machine learning	Machine learning
Oprindelsesår	2018–2020	2010 (formalized); 1990s (early roots)
Ophavsperson	Wu, Z. et al. / Chen, T. et al.	Pan, S. J. & Yang, Q. (survey); Bengio, Y. (deep learning framing)
Type	Self-supervised + non-parametric classifier	Learning paradigm
Oprindelig kilde	Chen, T., Kornblith, S., Norouzi, M., & Hinton, G. (2020). A simple framework for contrastive learning of visual representations. In Proceedings of the 37th International Conference on Machine Learning (ICML), PMLR 119, 1597–1607.	Pan, S. J., & Yang, Q. (2010). A Survey on Transfer Learning. IEEE Transactions on Knowledge and Data Engineering, 22(10), 1345–1359.
Aliasser	SSL-kNN, self-supervised kNN classifier, kNN evaluation probe, nearest-neighbor self-supervised classifier	TL, domain adaptation, fine-tuning, pre-trained model adaptation
Relaterede	4	3
Resumé	Self-supervised K-nearest neighbors (SSL-kNN) combines representation learning without labels with a non-parametric k-NN classifier. A neural encoder is first trained via a self-supervised objective — such as contrastive or masked prediction — so that semantically similar samples cluster together in the embedding space. A simple k-NN lookup on those embeddings then assigns class labels, serving both as a lightweight probe and as a practical classifier.	Transfer learning is a machine learning paradigm in which knowledge gained from training a model on a source task or domain is reused to improve learning on a different but related target task or domain. It is especially powerful when labeled data for the target task is scarce, and it underlies most modern deep learning applications in computer vision, natural language processing, and beyond.