%0 Journal Article %1 SMITH1980177 %A Smith, Stephen P. %A Dubes, Richard %D 1980 %J Pattern Recognition %K clustering methods paywall statistics visualization %N 3 %P 177 - 187 %R https://doi.org/10.1016/0031-3203(80)90042-4 %T Stability of a hierarchical clustering %U http://www.sciencedirect.com/science/article/pii/0031320380900424 %V 12 %X Clustering algorithms have the annoying habit of finding clusters even when the data are generated randomly. Verifying that potential clusterings are real in some objective sense is receiving more attention as the number of new clustering algorithms and their applications grow. We consider one aspect of this question and study the stability of a hierarchical structure with a variation on a measure of stability proposed in the literature.(1,2) Our measure of stability is appropriate for proximity matrices whose entries are on an ordinal scale. We randomly split the data set, cluster the two halves, and compare the two hierarchical clusterings with the clustering achieved on the entire data set. Two stability statistics, based on the Goodman-Kruskal rank correlation coefficient, are defined. The distributions of these statistics are estimated with Monte Carlo techniques for two clustering methods (single-link and complete-link) and under two conditions (randomly selected proximity matrices and proximity matrices with good hierarchical structure). The stability measures are applied to some real data sets.

@article{SMITH1980177, abstract = {Clustering algorithms have the annoying habit of finding clusters even when the data are generated randomly. Verifying that potential clusterings are real in some objective sense is receiving more attention as the number of new clustering algorithms and their applications grow. We consider one aspect of this question and study the stability of a hierarchical structure with a variation on a measure of stability proposed in the literature.(1,2) Our measure of stability is appropriate for proximity matrices whose entries are on an ordinal scale. We randomly split the data set, cluster the two halves, and compare the two hierarchical clusterings with the clustering achieved on the entire data set. Two stability statistics, based on the Goodman-Kruskal rank correlation coefficient, are defined. The distributions of these statistics are estimated with Monte Carlo techniques for two clustering methods (single-link and complete-link) and under two conditions (randomly selected proximity matrices and proximity matrices with good hierarchical structure). The stability measures are applied to some real data sets.}, added-at = {2019-12-18T23:29:28.000+0100}, author = {Smith, Stephen P. and Dubes, Richard}, biburl = {https://www.bibsonomy.org/bibtex/2a503a754014b636d633f8a2e4f2582ca/marcsaric}, description = {Stability of a hierarchical clustering - ScienceDirect}, doi = {https://doi.org/10.1016/0031-3203(80)90042-4}, interhash = {1a8644be55df93ce7fea970fd898d0c8}, intrahash = {a503a754014b636d633f8a2e4f2582ca}, issn = {0031-3203}, journal = {Pattern Recognition}, keywords = {clustering methods paywall statistics visualization}, number = 3, pages = {177 - 187}, timestamp = {2019-12-18T23:29:28.000+0100}, title = {Stability of a hierarchical clustering}, url = {http://www.sciencedirect.com/science/article/pii/0031320380900424}, volume = 12, year = 1980 }