方法对比

	自监督 DBSCAN	半监督DBSCAN
领域	机器学习	机器学习
方法族	Machine learning	Machine learning
起源年份	2018–2021	2000s
提出者	Ester et al. (DBSCAN base); pipeline pattern established in multiple works c. 2018–2021	Ester, M. et al. (DBSCAN base); semi-supervised extensions by multiple authors (2000s–2010s)
类型	Two-stage pipeline (self-supervised pre-training + density-based clustering)	Constrained density-based clustering
开创性文献	Ester, M., Kriegel, H.-P., Sander, J., & Xu, X. (1996). A density-based algorithm for discovering clusters in large spatial databases with noise. In Proceedings of the 2nd International Conference on Knowledge Discovery and Data Mining (KDD-96), pp. 226–231. AAAI Press.	Ester, M., Kriegel, H.-P., Sander, J., & Xu, X. (1996). A density-based algorithm for discovering clusters in large spatial databases with noise. In Proceedings of the 2nd International Conference on Knowledge Discovery and Data Mining (KDD-96), pp. 226–231. AAAI Press.
别名	SSL-DBSCAN, self-supervised density clustering, contrastive DBSCAN, representation-based DBSCAN	Constrained DBSCAN, SS-DBSCAN, DBSCAN with must-link/cannot-link constraints, seeded DBSCAN
相关	5	5
摘要	Self-supervised DBSCAN is a two-stage unsupervised pipeline that first trains a neural encoder on a pretext task — such as contrastive learning or masked reconstruction — to produce compact, semantically meaningful embeddings from unlabeled data, and then applies DBSCAN in the resulting embedding space to discover arbitrarily shaped clusters without requiring any class labels.	Semi-supervised DBSCAN extends the canonical density-based clustering algorithm (Ester et al., 1996) by incorporating a small set of pairwise or label constraints — must-link pairs that must share a cluster, cannot-link pairs that must be separated, or a handful of known labels — to guide cluster formation while retaining DBSCAN's ability to discover arbitrary-shaped clusters and flag noise points.