Statistical and Numerical Computing
Published Online: 15 JAN 2013
Copyright © 2002 John Wiley & Sons, Ltd
Encyclopedia of Environmetrics
How to Cite
Chenouri, S. 2013. Clustering . Encyclopedia of Environmetrics. 1.
- Published Online: 15 JAN 2013
Cluster analysis is concerned with the subdivision of individual objects into natural groups on the basis of collected data relating to these objects. The data may consist of observations on each object, or of perceived similarities or dissimilarities between pairs. As a statistical technique, cluster analysis is a form of “unsupervised classification,” as opposed to supervised classification, in which the groups and the group membership of the data are known, and are used to form rules to classify future observations. It is closely related to ordination, including principal component analysis and projection pursuit. These techniques are often used to justify the results of cluster analysis.
- gap statistic;
- hierarchical clustering;
- mixture models;
- similarity measures;
- spectral clustering