Semi-supervised Clustering by Seeding

By Sugato Basu, Arindam Banerjee, Raymond J. Mooney. International Conference on Machine Learning, 2002

1,010 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/f4f3a10d96e0b6d134e7e347e1727b7438d4006f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

Learning to Detect Patterns of Crime (2013, unknown): “, [18, 19]) is slightly related to our approach, in the sense that the set of patterns previously labeled by the police can be used as constraints for learned clusters; on the other hand, each of our clusters has different properties corresponding to different M.”
Deep k-Means: Re-Training and Parameter Sharing with Harder Cluster Assignments for Compressing Deep Convolutions (2018, background): “Without referring to sophisticated options such as semi-supervised clustering (Basu et al., 2002), we follow a simple heuristic which incurs almost no extra complexity: at each time of “lazy update” for layer W ∈ R s × N , we ﬁrst rank all N columns of W in terms of their (cid:96) 2 norms.”
A Survey of Text Classification Algorithms (2012, background): “This general approach is also referred to as co-training [9, 13, 37].”
A Survey of Text Clustering Algorithms (2012, unknown): “A similar approach has also been used in [15], except a wider variation of how the seeds may be selected has been explored.”
Inferring the outcomes of rejected loans: an application of semisupervised clustering (2019, methodology): “Basu et al. (2002) proposed two semisupervised K-means clustering methods: seeded K-means and constrained K-means. In seeded K-means, the seed data are used only for initialization but not in the subsequent clustering process, whereas, in constrained K-means, the labels of the non-seed data are re-e”
Integrating constraints and metric learning in semi-supervised clustering (2004, methodology): “It can be shown that the K-Means algorithm is essentially an EM algorithm on a mixture of K Gaussians under assumptions of identity covariance of the Gaussians, uniform mixture component priors and expectation under a particular type of conditional distribution (Basu et al., 2002).”
A probabilistic framework for semi-supervised clustering (2004, methodology): “It is well-known that KMeans is equivalent to an EM algorithm with hard clustering assignments [26, 6, 3].”
Enhancing Person Re-identification in a Self-Trained Subspace (2017, methodology): “In [45], this strategy is used for semi-supervised clustering.”
Constrained Semi-supervised Learning in the Presence of Unanticipated Classes (2016, methodology): “Basu and Mooney [9] proposed a seeded version of K-Means, which is very analogous to Nigam et al’s semi-supervised Naive Bayes, as another technique for semi-supervised learning.”
Building text classifiers using positive and unlabeled examples (2003, background): “Finally, our work is related to learning using a small labeled set and a large unlabeled set [2][3][4][5][10] [11][25][26][28][35].”
Topic-bridged PLSA for cross-domain text classification (2008, background): “Semi-supervised clustering [1] builds clusters under some additional constraints provided by a few labeled data, in the form of must-links (two examples must in the same cluster) and cannot-links (two examples cannot in the same cluster).”
CECM: Constrained evidential C-means algorithm (2012, methodology): “Given a mass function m, it is possible to define a plausibility function pl : 2 → [0, 1] and a belief function bel : 2 → [0, 1] by: pl(A) = ∑”
Semi-supervised graph clustering: a kernel approach (2005, background): “Research on semi-supervised clustering has considered supervision in the form of both labeled points (Demiriz et al. 1999; Sinkkonen and Kaski 2002; Basu et al. 2002) or constraints (Wagstaff et al. 2001; Klein et al. 2002; Xing et al. 2003; Bar-Hillel et al. 2003; Bie et al. 2003; Kamvar et al.…”