Clustering: Science or Art?

Ulrike von Luxburg, Robert C. Williamson, Isabelle Guyon
Proceedings of ICML Workshop on Unsupervised and Transfer Learning, PMLR 27:65-79, 2012.

Abstract

We examine whether the quality of different clustering algorithms can be compared by a general, scientifically sound procedure which is independent of particular clustering algorithms. We argue that the major obstacle is the difficulty in evaluating a clustering algorithm without taking into account the context: why does the user cluster his data in the first place, and what does he want to do with the clustering afterwards? We argue that clustering should not be treated as an application-independent mathematical problem, but should always be studied in the context of its end-use. Different techniques to evaluate clustering algorithms have to be developed for different uses of clustering. To simplify this procedure we argue that it will be useful to build a “taxonomy of clustering problems” to identify clustering applications which can be treated in a unified way and that such an effort will be more fruitful than attempting the impossible – developing “optimal” domain-independent clustering algorithms or even classifying clustering algorithms in terms of how they work.

Cite this Paper


BibTeX
@InProceedings{pmlr-v27-luxburg12a, title = {Clustering: Science or Art?}, author = {von Luxburg, Ulrike and Williamson, Robert C. and Guyon, Isabelle}, booktitle = {Proceedings of ICML Workshop on Unsupervised and Transfer Learning}, pages = {65--79}, year = {2012}, editor = {Guyon, Isabelle and Dror, Gideon and Lemaire, Vincent and Taylor, Graham and Silver, Daniel}, volume = {27}, series = {Proceedings of Machine Learning Research}, address = {Bellevue, Washington, USA}, month = {02 Jul}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v27/luxburg12a/luxburg12a.pdf}, url = {https://proceedings.mlr.press/v27/luxburg12a.html}, abstract = {We examine whether the quality of different clustering algorithms can be compared by a general, scientifically sound procedure which is independent of particular clustering algorithms. We argue that the major obstacle is the difficulty in evaluating a clustering algorithm without taking into account the context: why does the user cluster his data in the first place, and what does he want to do with the clustering afterwards? We argue that clustering should not be treated as an application-independent mathematical problem, but should always be studied in the context of its end-use. Different techniques to evaluate clustering algorithms have to be developed for different uses of clustering. To simplify this procedure we argue that it will be useful to build a “taxonomy of clustering problems” to identify clustering applications which can be treated in a unified way and that such an effort will be more fruitful than attempting the impossible – developing “optimal” domain-independent clustering algorithms or even classifying clustering algorithms in terms of how they work. } }
Endnote
%0 Conference Paper %T Clustering: Science or Art? %A Ulrike von Luxburg %A Robert C. Williamson %A Isabelle Guyon %B Proceedings of ICML Workshop on Unsupervised and Transfer Learning %C Proceedings of Machine Learning Research %D 2012 %E Isabelle Guyon %E Gideon Dror %E Vincent Lemaire %E Graham Taylor %E Daniel Silver %F pmlr-v27-luxburg12a %I PMLR %P 65--79 %U https://proceedings.mlr.press/v27/luxburg12a.html %V 27 %X We examine whether the quality of different clustering algorithms can be compared by a general, scientifically sound procedure which is independent of particular clustering algorithms. We argue that the major obstacle is the difficulty in evaluating a clustering algorithm without taking into account the context: why does the user cluster his data in the first place, and what does he want to do with the clustering afterwards? We argue that clustering should not be treated as an application-independent mathematical problem, but should always be studied in the context of its end-use. Different techniques to evaluate clustering algorithms have to be developed for different uses of clustering. To simplify this procedure we argue that it will be useful to build a “taxonomy of clustering problems” to identify clustering applications which can be treated in a unified way and that such an effort will be more fruitful than attempting the impossible – developing “optimal” domain-independent clustering algorithms or even classifying clustering algorithms in terms of how they work.
RIS
TY - CPAPER TI - Clustering: Science or Art? AU - Ulrike von Luxburg AU - Robert C. Williamson AU - Isabelle Guyon BT - Proceedings of ICML Workshop on Unsupervised and Transfer Learning DA - 2012/06/27 ED - Isabelle Guyon ED - Gideon Dror ED - Vincent Lemaire ED - Graham Taylor ED - Daniel Silver ID - pmlr-v27-luxburg12a PB - PMLR DP - Proceedings of Machine Learning Research VL - 27 SP - 65 EP - 79 L1 - http://proceedings.mlr.press/v27/luxburg12a/luxburg12a.pdf UR - https://proceedings.mlr.press/v27/luxburg12a.html AB - We examine whether the quality of different clustering algorithms can be compared by a general, scientifically sound procedure which is independent of particular clustering algorithms. We argue that the major obstacle is the difficulty in evaluating a clustering algorithm without taking into account the context: why does the user cluster his data in the first place, and what does he want to do with the clustering afterwards? We argue that clustering should not be treated as an application-independent mathematical problem, but should always be studied in the context of its end-use. Different techniques to evaluate clustering algorithms have to be developed for different uses of clustering. To simplify this procedure we argue that it will be useful to build a “taxonomy of clustering problems” to identify clustering applications which can be treated in a unified way and that such an effort will be more fruitful than attempting the impossible – developing “optimal” domain-independent clustering algorithms or even classifying clustering algorithms in terms of how they work. ER -
APA
von Luxburg, U., Williamson, R.C. & Guyon, I.. (2012). Clustering: Science or Art?. Proceedings of ICML Workshop on Unsupervised and Transfer Learning, in Proceedings of Machine Learning Research 27:65-79 Available from https://proceedings.mlr.press/v27/luxburg12a.html.

Related Material