Clustering: Science or Art?

By Ulrike von Luxburg, Robert C. Williamson, I. Guyon. ICML Unsupervised and Transfer Learning, 2009

347 citations recorded by the source corpus.

View Paper: https://www.semanticscholar.org/paper/e46e683a1167990f8778335b4536c27b624be0cf

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

K-means properties on six clustering benchmark datasets (2018, background): “[12] discuss the pros and cons of how to choose benchmarking data.”
Functional data clustering: a survey (2013, methodology): “For instance, techniques using simulations or classiﬁcation datasets are both questionable (see for instance Guyon et al. (2009) for an interesting discussion on this topic).”
PAC-Bayesian Analysis of Co-clustering and Beyond (2010, background): “The need to consider unsupervised learning within the context of its subsequent application has been pointed out by many researchers, especially those concerned with practical applications of these methods (Guyon et al., 2009).”
A Hierarchy of Limitations in Machine Learning (2020, background): “It might be associated with claims about uncovering underlying structure, but there is ample work (von Luxburg et al., 2012) about the difficulty of validating clusters beyond the face validity of “making sense.””
Robust clustering by detecting density peaks and assigning points based on fuzzy weighted K-nearest neighbors (2016, background): “Clustering is important for un3 covering the inherent, potential and unknown knowledge, principles or rules in the real-world, and has been widely used 4 in scientific and engineering applications [12,15,18,27,30,33] .”
Simultaneously Discovering and Quantifying Risk Types from Textual Risk Disclosures (2014, background): “Von Luxburg et al. (2012) suggest that unsupervised clustering methods should always be studied and evaluated in the context of their end use.”