Dendrogrammen

A dendrogram is a tree diagram that illustrates the arrangement of the clusters produced by hierarchical clustering. The word comes from the Greek “dendron” meaning tree and the German plural “Dendrogrammen” refers to multiple such diagrams. In a dendrogram, the horizontal axis is usually the number of observations, while the vertical axis represents the dissimilarity or distance at which clusters merge. Each leaf corresponds to an individual data point, and branches join points or groups that are most similar according to a chosen metric.

Hierarchical clustering can be agglomerative, starting with each observation as its own cluster and successively merging the closest pairs, or divisive, beginning with all observations in one cluster and breaking them apart. The resulting dendrogram records the sequence of merges or splits, allowing researchers to decide on a meaningful number of clusters by cutting the tree at a chosen dissimilarity level. The choice of distance metric (e.g., Euclidean, Manhattan) and linkage criterion (single, complete, average, Ward’s) strongly influences the shape of the dendrogram and the resulting cluster structure.

Dendrogrammen are widely used in biology to depict evolutionary relationships (phylogenetic trees), in text mining to group documents, in market research to segment consumer segments, and in machine learning for exploratory data analysis. Common software packages that generate dendrograms include R (the hclust function), Python’s scipy.cluster.hierarchy module, and specialized bioinformatics tools such as MEGA and PhylIP. Interpretation of a dendrogram requires careful attention to the scale of the vertical axis, as different axes may represent raw dissimilarity values, distances, or probabilities, and to the potential presence of outliers that can distort the overall structure.