Article

Classification Analysis

Otto Wildi

in Ecology

ISBN: 9780199830060
Published online May 2012 | | DOI: http://dx.doi.org/10.1093/obo/9780199830060-0025
Classification Analysis

More Like This

Show all results sharing these subjects:

  • Applied Ecology (Environmental Science)
  • Ecology and Conservation
  • Plant Ecology
  • Zoology and Animal Sciences

GO

Show Summary Details

Preview

Classification is the grouping of similar objects to produce a classification. It is a necessary process of all living creatures to maintain an overview of the complex environment around them by reducing the amount of information captured by perception (see Everitt, et al. 2011, Cluster Analysis, cited under Clustering and Classification). Modern biology and as well as ecology could evolve only when a universal classification of all living beings was finally possible, first proposed by Carolus Linnaeus in 1735. This article focuses on the relatively recent development of computerized methods that help to automatically classify information into large and complex datasets. As the section Clustering and Classification explains in more detail, this is elusive in many ways. First, all classifications serve specific purposes. Multipurpose investigations hence may require alternative classifications of the same body of data. Second, even if criteria to measure success of classification are set, modern computers still lack the ability to find the optimum solution not only for large but also for small datasets, and most algorithms are merely attempts to approximate these. Although there have been phases when classification analysis played a prominent role in various fields of natural science, it must be seen as a branch of mathematics or statistics. Everitt, et al 2011 (see Clustering and Classification), for example, presents applications in market research, astronomy, psychiatry, weather forecasting, archaeology, bioinformatics, and genetics, and Gan, et al. 2007 (cited under Clustering and Classification) concentrates on gene expression. Whereas trends in mathematics are definitely not a focus of this article, the role of classification in ecology in the past, present, and future is evaluated. One conclusion is that there are hardly any ecological journals that specialize in classification, nor are there any in mathematics. A list of seventy-six journals in which articles on clustering are regularly published is given in the Data Clustering section; this is an impressive number at first glance. There is even a Journal of Classification. However, this publication covers the field of multivariate analysis in general, with rather diverse applications and no recent contribution to ecology. A search in the ISI Web of Science using key words such as “classification,” “clustering,” “cluster analysis,” “group structure,” “numerical ecology,” and “numerical taxonomy,” in combination with “animal ecology,” “plant ecology,” “marine ecology,” “functional traits,” and others delivers a different picture. As can be expected, classification is omnipresent in ecology, but unlike other statistical models it plays a subordinate role. There are hardly any papers in which cluster analysis is the ultimate method used; in almost all cases, it is involved only in initial data screening if even considered at all. The field in which it is most frequently used is (plant) community ecology. As is shown in this article, animal ecology and—even more so—marine ecology inherited many of these methods.

Article.  8539 words. 

Subjects: Applied Ecology (Environmental Science) ; Ecology and Conservation ; Plant Ecology ; Zoology and Animal Sciences

Full text: subscription required

How to subscribe Recommend to my Librarian

Buy this work at Oxford University Press »

Users without a subscription are not able to see the full content. Please, subscribe or login to access all content.