So sánh phương pháp

	Học độ đo bán giám sát	Học tăng cường tự giám sát
Lĩnh vực	Học máy	Học máy
Họ	Machine learning	Machine learning
Năm ra đời	2007–2008	2018–2020
Người khởi xướng	Yeung, D.-Y. & Chang, H.; Davis, J. V. & Dhillon, I. S.	LeCun, Y. and community (formalized ~2018–2020)
Loại	Hybrid supervised/unsupervised distance learning	Representation learning paradigm
Công trình gốc	Yeung, D.-Y., & Chang, H. (2007). A kernel approach for semi-supervised metric learning. IEEE Transactions on Neural Networks, 18(1), 141–149.	LeCun, Y. & Misra, I. (2022). Self-supervised learning: The dark matter of intelligence. Meta AI Blog. https://ai.facebook.com/blog/self-supervised-learning-the-dark-matter-of-intelligence/
Tên gọi khác	SSML, semi-supervised distance learning, constrained metric learning, weakly supervised metric learning	SSL, self-supervised pre-training, pretext-task learning, unsupervised representation learning
Liên quan	5	3
Tóm tắt	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.	Self-supervised learning (SSL) is a machine-learning paradigm that generates its own supervisory signal directly from unlabeled data by defining an auxiliary pretext task — such as predicting masked words, rotating images, or contrasting augmented views — and uses the learned representations as a powerful starting point for downstream tasks with minimal labeled examples.