Cluster Analysis is one of the most favorite multivariable statistical methods, which is actually a special type of aggregating method. Observations are clustered by variables belonged to the observations. Our purpose is to create such clusters, in which the elements are the most similar, and between the clusters they are the most variant. For example these clusters could be the qualitative classifications of farms.
There have been several methods in Cluster Analysis as well as numerous distance measures, which could be used. In this article, we study all of these methods and measures. After we show the theoretical background, we apply the method in a given casestudy to control the qualitative classifications of experts. In this study, we use both the hierarchical and the non-hierarchical method, and also compare them. We would like to attract the attention that the most important problem of the analysis is to determine the optimal of clusters.